WO1997031247A1 - Computer color matching method and apparatus - Google Patents

Computer color matching method and apparatus Download PDF

Info

Publication number
WO1997031247A1
WO1997031247A1 PCT/JP1996/000738 JP9600738W WO9731247A1 WO 1997031247 A1 WO1997031247 A1 WO 1997031247A1 JP 9600738 W JP9600738 W JP 9600738W WO 9731247 A1 WO9731247 A1 WO 9731247A1
Authority
WO
WIPO (PCT)
Prior art keywords
color
colorant
mixture
value
computer
Prior art date
Application number
PCT/JP1996/000738
Other languages
French (fr)
Japanese (ja)
Other versions
WO1997031247A6 (en
Inventor
Hiroshi Kumamoto
Hideharu Imoto
Hiroshi Tamae
Nobuo Adachi
Original Assignee
Toto Ltd
Hiroshi Kumamoto
Hideharu Imoto
Hiroshi Tamae
Nobuo Adachi
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Toto Ltd, Hiroshi Kumamoto, Hideharu Imoto, Hiroshi Tamae, Nobuo Adachi filed Critical Toto Ltd
Priority to JP52996597A priority Critical patent/JP3870421B2/en
Publication of WO1997031247A1 publication Critical patent/WO1997031247A1/en
Publication of WO1997031247A6 publication Critical patent/WO1997031247A6/en

Links

Classifications

    • GPHYSICS
    • G01MEASURING; TESTING
    • G01JMEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY
    • G01J3/00Spectrometry; Spectrophotometry; Monochromators; Measuring colours
    • G01J3/46Measurement of colour; Colour measuring devices, e.g. colorimeters
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01JMEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY
    • G01J3/00Spectrometry; Spectrophotometry; Monochromators; Measuring colours
    • G01J3/46Measurement of colour; Colour measuring devices, e.g. colorimeters
    • G01J3/463Colour matching

Definitions

  • the present invention relates to a method and apparatus for predicting the mixing ratio of a colorant or the color of a mixture by computer color matching.
  • the tristimulus values X, ⁇ , and ⁇ of the mixture are calculated. Therefore, the color of the mixture can be calculated.
  • the absorption coefficient ( ⁇ ) and scattering coefficient S ⁇ ( ⁇ ) of the material to be colored and various colorants must be determined in advance.
  • the absolute method is a method for calculating the absolute values of the absorption coefficient ⁇ ⁇ ( ⁇ ) and the scattering coefficient (;.) Of each substance.
  • the relative method a primary pigment (usually white pigment) assuming scattering coefficient s w 1 and the absorption coefficient of each substance
  • Equation 2 This is a method of calculating the relative value of ( ⁇ ) and the scattering coefficient Si (A).
  • the spectral reflectance R) of the mixture is given by the ratio of the absorption coefficient K M and the scattering coefficient S M of the mixture. Therefore, even if the absolute values of the absorption coefficient Ki (;.) And the scattering coefficient Si ( ⁇ ) of each substance are unknown, if the relative values of the absorption coefficient ( ⁇ ) and the scattering coefficient Si ( ⁇ ) of each substance are known, From Equations 1 and 2, the correct spectral reflectance R ( ⁇ ) can be determined. Since the preparation of samples for the absolute method is quite difficult and requires complicated work, the relative method is usually used.
  • the absorption coefficient K w ( ⁇ ) of the reference white pigment is calculated as follows. to determine the scattering coefficient S w. First, a mixture is prepared by mixing only a white pigment with an object to be colored, and its spectral reflectance R ( ⁇ ) is measured. In the conventional relative method, the object to be colored is regarded as colorless and transparent, and it is assumed that both the absorption coefficient and the scattering coefficient of the object to be colored are zero.
  • the absorption coefficient given by Equation 1 above kappa Micromax (lambda) and the scattering coefficient S “(lambda) is the scattering coefficient of the white pigment S w and absorption coefficient K w ( ⁇ ) equal to, respectively.
  • an object to be colored is regarded as colorless and transparent, and its absorption coefficient and scattering coefficient are assumed to be zero.
  • the objects to be colored Glaze layers in the case of ceramics
  • the blending ratio of the white pigment is large, the error caused by assuming that the object to be colored is colorless and transparent is small, but when the blending ratio of the white pigment is small, the error becomes so large that it cannot be ignored. In order to avoid such errors, it is necessary to take into account the absorption coefficient and scattering coefficient of the colorless and non-transparent object in Equation 1 above.
  • a first object of the present invention is to perform computer color matching in consideration of an absorption coefficient and a scattering coefficient of an object to be colored that is not colorless and transparent.
  • a second object of the present invention is to reduce a prediction error in computer color matching.
  • a third object of the present invention is to make effective use of the prepared colorant by using computer color matching and to simplify re-formulation.
  • the present invention relates to a method for predicting a mixing ratio of a colorant or a color of a mixture by computer color matching.
  • This computer color matching method is
  • the basis of the scattering coefficient S w ' a step of determining a format which depends the turbulent coefficient dispersion and absorption coefficient K p of the colored colorant not white S p in Formulation ratio C p of the colored colorant, a desired color Mixing ratio of the colorant to adjust the mixture having The color of the mixture produced in formulation ratio, the absorption coefficient K w ', p, and the scattering coefficient S w', a step of determining by performing computer color matching using S p,
  • the scattering coefficient s w of the second mixture obtained by mixing the white colorant with the non-colorless and transparent coloring object is expressed by a function f (C w ), and this scattering coefficient s w , is used as a reference.
  • the scattering coefficient S vom′ of the plurality of the second mixtures is represented by a function f (C w ) of the mixing ratio C w of the white colorant, and the coefficient f (C w ) Temporarily determining the value of
  • the coefficients of the function f (C w ) can be obtained so that the prediction accuracy by computer color matching is improved.
  • a computer color matching method includes: (a) mixing a plurality of colorants to prepare a plurality of samples having different mixing ratios; and (b) the plurality of samples. Measuring the spectral reflectances of the plurality of samples, and obtaining actual measured values of coordinate values of a predetermined color system representing respective colors of the plurality of samples from the measured values of the spectral reflectances; Calculating, for each of the samples, a prediction error of the coordinate values of the color system; and Analyzing the relationship between the coordinate values of the color system with respect to the sample and the prediction error by a predetermined error correction method; and (e) calculating a target value or a prediction value of computer color matching using the error correction method. Predicting the blending ratio of the colorant in the new mixture or predicting the color of the mixture by computer color matching while making corrections.
  • the coordinate values of the predetermined color system and the prediction errors thereof for a plurality of samples are analyzed by a predetermined error correction method, and the prediction is performed while correcting the target value or the predicted value of the computer color matching by the error correction method. it is possible to reduce the prediction error without correcting the absorption coefficient Ki and scattering coefficient number S Alpha of each component.
  • the step (d) includes a step of causing a dual neural network to learn a relationship between the coordinate values of the color system and the prediction error for the plurality of samples, and the step (e). ) Shows a process of performing prediction by computer color matching using a trained neural network.
  • the neural network has a three-layer hierarchical structure including an input layer including three neurons, an intermediate layer including a plurality of neurons, and an output layer including three neurons. It is preferred to have.
  • a computer color matching method is a method for predicting a mixing ratio of a colorant in a target mixture having a desired color
  • the target colorimetric coordinate of the target mixture is corrected by the calculation error of the coordinate value of the colorimetric system of the adjacent color sample, and computer color matching is performed using the corrected target value.
  • Computer color matching can be performed under conditions that reduce value calculation errors.
  • a color difference from the target mixture is minimized from a database including a mixing ratio of a coloring agent and an actual measurement value of coordinates of the color system.
  • the step of searching for the adjacent color sample by selecting the sample to be performed is provided.
  • a computer color matching method includes: adjusting a compounding ratio of a plurality of colorants so that the mixture obtained by mixing the plurality of colorants exhibits a color similar to a desired target color. Is the method of seeking
  • the amount of change in the calculated color evaluation value at that time is determined. Then, based on the amount of change of the color evaluation value calculation amount, the amount of increase correction for each colorant is calculated, and the difference between the actually measured value of the color evaluation value of the preparation sample and that of the primary preparation is determined within a predetermined range. To match. For this reason, if each colorant is actually added to the primary formulation by the calculated halo correction amount, the primary formulation will be added to the target color of the sample of the formulation or a formulation exhibiting a color similar to the target color. Can be reconstituted.
  • the computer color matching method of the present invention there is no need to obtain a negative correction amount meaning removal of the colorant, so that it is not necessary to dispose of the preparation in which the colorant has been prepared (primary preparation).
  • the existing formulation can be used effectively. In this case, even if the primary formulation is newly prepared on a trial basis instead of the existing formulation, the technician need only be involved once in the formulation, and at that time, Since no intuition or experience is required, re-mixing of the composition can be simplified.
  • the color evaluation value is calculated based on the assumption that a small amount of the coloring agent is corrected to be added to the primary formulation compared to the blending amount of the coloring agent in the secondary formulation. Calculating the amount of change in the calculated value.
  • a temporary increase correction of the colorant which is performed in order to obtain a change amount of the calculated value of the color evaluation value, is performed in a minute unit, and a change amount in a minute unit can be obtained.
  • the step (e) includes a step of calculating the minimum amount of increase correction for each of the colorants by a linear measurement surface method using a cost function representing a derivation cost associated with the increase in the amount of the colorant. It is preferable to settle.
  • the present invention is also directed to an apparatus for predicting a mixing ratio of a colorant or a color of a mixture by computer color matching.
  • This device has a scattering coefficient S w _ of a second mixture in which a white colorant is mixed with a colorless and transparent object to be colored.
  • the absorption coefficient K P and the scattering coefficient s P of the non-white colored colorant are determined in a form depending on the mixing ratio c p of the colored colorant, and a desired color is obtained.
  • the blending ratio of a young agent for preparing a mixture having the following formulas, or the color of the mixture produced at a predetermined blending ratio, is calculated using the absorption coefficients K w ′, K P and the scattering coefficients s w ′, S P Means to obtain by performing the computer color matching used;
  • a computer color matching device comprises: means for measuring spectral reflectances of a plurality of samples having different mixing ratios formed by mixing a plurality of colorants;
  • a computer color matching apparatus includes: an actual measurement value of coordinates of a predetermined color system for a proximity color sample having a colorant mixing ratio known and having a color close to the desired color; Means to determine
  • a computer color matching device calculates a blending ratio of a plurality of coloring agents so that the formulation obtained by blending the plurality of coloring agents exhibits a color similar to a desired target color. Things.
  • Each of the colorants based on the amount of change in the calculated value of the color evaluation value so that the difference between the measured value of the color evaluation value of the preparation sample and that of the primary preparation matches within a predetermined range.
  • Means for calculating the increase correction amount Means for calculating the increase correction amount.
  • the invention is further directed to a method for determining the absorption and scattering coefficients of a colorant used in computer color matching. This method
  • the scattering coefficient s w ′ of the second mixture is given by the following relationship involving the mixing ratio c w of the white colorant, a constant 3 l and a coefficient S B :
  • the scattering coefficient S w 'of the second mixture, the mixing ratio C w of the white colorant and the following relations a, a, b, d, e, C WU Given by:
  • the present invention is also directed to mixed glazes made using computer color matching. This mixed glaze
  • the relative values of the absorption coefficient K p and the scattering coefficient S p of the non-white colored colorant are determined, and the absorption coefficient K w , , K P and the scattering coefficient s w,, by performing Konbiyu Takara over matching using S P, formulation ratio of the colorant to adjust the mixed glaze having a desired color, or, at a predetermined compounding ratio Predict the color of the resulting mixed glaze,
  • the present invention is further directed to ceramics manufactured using a mixed glaze made using computer color matching. This ceramic is
  • the relative values of the absorption coefficient ⁇ ⁇ and the scattering coefficient S p of the non-white colored colorant are determined, and the absorption coefficient K w , , K p and the scattering coefficient s w, by performing Konbiyu Takara over pine quenching with S p, formulation ratio of the colorant to adjust the mixed glaze having a desired color or a predetermined compounding ratio Predict the color of the mixed glaze generated by
  • a method for determining an absorption coefficient and a scattering coefficient of a colorant used in computer color matching comprises:
  • an apparatus for determining an absorption coefficient and a scattering coefficient of a colorant used in computer color matching comprises:
  • the present invention is also directed to a toilet bowl manufactured using mixed glaze made using computer color matching.
  • FIG. 1 is a flowchart showing the overall processing procedure in the first embodiment.
  • Fig. 2 is a graph showing the scattering coefficient S w 'given by Eq.
  • Fig. 3 is a flowchart showing the detailed procedure of step S1.
  • Fig. 4 is an explanatory diagram showing the mixing ratio of a sample to determine the physical properties of (base glaze + white pigment).
  • Figure 5 is a conceptual diagram showing the spectral reflectance R (;.) Of the sample.
  • FIG. 6 is a flowchart showing the detailed procedure of step S14.
  • Fig. 7 is an explanatory diagram showing the mixing ratio of the pigment physical property value determination sample.
  • FIG. 8 is an explanatory diagram showing the mixing ratio of the verification sample used in the first embodiment.
  • Figure 9 is a graph showing an example of the dependence of the absorption coefficient ⁇ ⁇ ( ⁇ ) on the blending rate C p .
  • FIG. 10 is an explanatory diagram showing a prediction result of a mixing ratio in the first embodiment.
  • FIG. 11 is a flowchart illustrating an overall procedure of a process according to the second embodiment.
  • Figure 12 is an explanatory diagram showing the configuration of a neural network.
  • FIG. 13 is a flowchart showing the detailed procedure of step S31.
  • Fig. 14 is a conceptual diagram showing the distribution of tristimulus values in the prediction target range PA and multiple samples M1 to M7 in computer color matching.
  • FIG. 15 is a conceptual diagram showing prediction errors of seven samples M1 to M7 in the first embodiment.
  • Figure 16 is an explanatory diagram showing the target tristimulus values (the values determined in step S41) and the prediction error AMi ( ⁇ , ⁇ , ⁇ ) for each sample.
  • Figure 17 is an explanatory diagram showing the results of the neural network learning in the second embodiment.
  • Fig. 18 is an explanatory diagram showing the verification results of the prediction of computer color matching in the second embodiment.
  • Fig. 19 is an X-Y chromaticity diagram showing the variation between the designed colors (standard colors) and the colors of the ceramics actually manufactured.
  • FIG. 20 is a ⁇ -chart showing the overall procedure of the process in the third embodiment.
  • FIG. 21 is a flowchart showing the detailed procedure of computer color matching in step S57.
  • FIG. 22 is a table showing the tristimulus values of the reference color sample and the tristimulus values of the overflow limit sample used in the third embodiment.
  • Fig. 23 is a table showing the results of the prediction of tristimulus values of the light limit sample, comparative examples, and the actual mixing ratio according to the third embodiment.
  • Figure 24 is a block diagram showing the equipment for implementing the computer color matching method of each embodiment.
  • FIG. 25 is a flowchart showing the overall procedure of the process in the computer color matching method of the fourth embodiment.
  • FIG. 26 is a flowchart showing the detailed processing of step S76 in FIG.
  • Figure 27 shows the comparison between the extreme U value (color value) for the target color cast obtained in step S72 and the value of the tristimulus for the first prototype glaze obtained in step S73, and Table showing the mixing ratio (mixing ratio) of each pigment in the target color sample glaze and the first prototype glaze.
  • Fig. 28 is a table showing the rate of change (derivative coefficient) of tristimulus values when each pigment is added to the first trial glaze shown in Fig. 27 by a small amount of each pigment.
  • FIG. 29 is a table showing the result of the computer color matching method according to the fourth embodiment.
  • FIG. 30 is a table showing another result according to the fourth embodiment.
  • FIG. 1 is a flowchart showing the overall procedure of the process in the embodiment.
  • the mixture of interest in the first embodiment is a glaze for covering the surface of the ceramic body.
  • the base glaze (base glaze) containing no pigment is the object to be colored, and the glaze obtained by adding a pigment to this base glaze is a mixture to be subjected to computer color matching.
  • an emulsifier is effective as a white pigment.
  • Emulsifiers include zirconium compounds such as zircon and phosphorus compounds such as calcium phosphate.
  • step S ⁇ the absorption coefficient K w _ 'and the scattering coefficient S w ' of the mixture of the object to be colored (base glaze) and the white pigment are determined.
  • the difference from the conventional relative method is that the material to be colored is not assumed to be colorless and transparent, and the absorption coefficient of the mixture is determined without calculating the absorption coefficient or scattering coefficient of the material to be colored or the white pigment alone.
  • the point is to find K w 'and the scattering coefficient s w '.
  • the scattering coefficient S w ' is a function depending on the compounding ratio C w of the white pigment f (C w
  • the absorption coefficient ⁇ ses' is determined in a form that depends on the white pigment blending ratio c w .
  • step S 2 pigment other than white pigment (hereinafter, “colored pigments” and hump) absorption coefficient K p and scattering coefficient S p of is also determined (step S 2).
  • the absorption and scattering coefficients of the chromatic color pigment is determined Me in a format depending on the compounding ratio c p.
  • step S3 computer color matching is performed using the above-described equations 1 and 2 based on the absorption coefficient and the scattering coefficient obtained in steps S1 and S2, and the color prediction of the mixture and the prediction of the blending ratio are performed. Do.
  • Equation 1 Equation 3
  • K w,, S w ' is the absorption coefficient and scattering coefficient of the mixture
  • K w, S w is the absorption coefficient and scattering coefficient of the white pigment alone
  • K B, S B is the scattering and absorption coefficient of the base glaze alone coefficient
  • C w is the mixing ratio of the white pigment
  • c B is the mixing ratio of the base glaze. It should be noted that the mixing ratios c w and c B should be expressed accurately by volume ratio, but even if expressed by weight ratio, the error is usually negligible.
  • Equation 4 the scattering coefficient S w 'of the mixture is given by Equation 4 below. , _ Cw S W + Les RSR. ⁇ ,
  • the scattering coefficient s w is independent of the mixing ratio c w of the white pigment.
  • the scattering coefficient S w for the mixture of (base glaze + white pigment) is determined. Then, if the scattering coefficient S w , is obtained, the absorption coefficient K w ′ can be obtained from Equation 2 using the measured value of the spectral reflectance R (A).
  • FIG. 3 is a flowchart showing a detailed procedure of step S1 in FIG.
  • step S 1 ⁇ a mixture (first mixture) was prepared in which only the base glaze and the white pigment were mixed.
  • FIG. 4 shows the mixing ratios of 13 samples W] 2 to W0 prepared in the first embodiment. As can be seen from FIG. 4, 1 3 samples Wl 2 ⁇ W0 are those created by the compounding ratio C w of the white pigment is changed by 1% in the range of 1 2% 0%.
  • the sample in the first example was prepared by baking a normal ceramic body with a blended glaze. In addition, other samples described later were prepared under the same conditions.
  • step S12 the spectral reflectance R '(A) of each sample W2 to W0 was measured with a spectrophotometer.
  • step S13 the measured values of spectral reflectance R,
  • Equation 6 Based on ( ⁇ ), the coefficients a and b in Equation 6 were experimentally determined as follows.
  • the coefficient, k 2 is a value that depends on the optical properties of the object to be colored (base glaze).
  • the coefficients, k 2 can be determined from the refractive index n of the object to be colored according to the following equation 10:
  • the base glaze used in the first embodiment has a refractive index ⁇ of about 1.4.
  • (KZS) wi is obtained from the above-described equation (2).
  • Equation 9 The calculation to determine the ideal state spectral reflectance R (1) from the spectral reflectance measured value R 'U) according to Equation 9 is performed when calculating (K / S) from the spectral reflectance of the sample in another process described later. Is similarly performed.
  • the base glaze mixed with a high concentration of white pigment exhibits almost constant reflectance over the entire visible wavelength range (about 400 nm to about 700 nm).
  • the scattering coefficient S wi ′ can be obtained from the value of the spectral reflectance R ( ⁇ ) using Expressions 2 and 8.
  • the scattering coefficient S has the same value in the entire wavelength range of visible light.
  • Expression 6 was rewritten into Expression 11 below.
  • step S14 of FIG. 3 the values of the coefficients d and e of Equation 7 on the low concentration side are determined.
  • the blending ratio C w of the white pigment is relatively small, the influence of the ground becomes large, so that the measured value of the spectroscopic reflectance R ( ⁇ ) is not information of only the true glaze layer. Therefore, it is not easy to determine the coefficients d and e based on the measured values of the spectral reflectance R ( ⁇ ). Therefore, the values of the coefficients d and e are determined according to the procedure shown in Fig. 6.
  • FIG. 7 is an explanatory diagram showing the mixing ratio of the pigment determination sample.
  • the sample for determining the physical properties of the pigment was prepared by mixing a base sleeve, a white pigment and another colored pigment.
  • the total compounding ratio of the white pigment and colored pigment (Pigment Volume Concentration, P VC) 1 2% - a constant, 1 formulatory ratio C p of colored pigment in the range of 1% to 1 2% 12 samples M] to M12 were prepared.
  • P VC ment Volume Concentration
  • sample Ml2 is a mixture containing no white pigment, and corresponds to the second mixture of the present invention.
  • Samples M1 to M11 correspond to the third mixture of the present invention. Hit.
  • step S22 a verification sample is created.
  • 8 are explanatory diagrams showing the mixing ratios of the verification samples used in the first example.
  • the verification samples D ⁇ to D4 correspond to the fourth mixture in the present invention.
  • step S23 temporary values are assigned to the coefficients d and e.
  • the coefficient d is also a rough value. It is possible to decide.
  • the value of the scattering coefficient S wi 'does not depend on the wavelength, but the value of the absorption coefficient K w ,' depends on the wavelength.
  • the scattering coefficient s wi ′ is used as a reference value that does not depend on the wavelength, and the relative values of other physical properties are obtained. Since the value of the spectral reflectance RU) depends of course on the wavelength, the value of (KZS) wi obtained according to Equation 2 also depends on the wavelength. Therefore, the absorption coefficient K wi 'obtained according to Equation 12 also depends on the wavelength. In other words, the absorption coefficient K wi 'is obtained in a form that depends on the blending ratio C w and the wavelength of the white pigment.
  • step S25 the absorption coefficient ⁇ ⁇ ( ⁇ ) and the scattering coefficient S p ( ⁇ ) of each colored pigment are determined using the pigment property value determination sample.
  • Step S25 corresponds to step S2 in FIG. (.)
  • SP (s) are calculated by the following procedure. First, for a pigment physical property value determination sample, the following expression 13 is obtained from expression 1 described above.
  • Equation 16 (KZS) p is the ratio between the absorption coefficient and the scattering coefficient of the colored pigment alone, (K / S) is the ratio between the absorption coefficient and the scattering coefficient of the pigment physical property determination sample, and (KZ S) is ( This is the ratio between the absorption coefficient and the scattering coefficient of the sample (base glaze + white pigment).
  • C w is the sum of the blending ratios of the base glaze and the white pigment, and s is the scattering coefficient of the mixture of (base glaze + white pigment), which is given by Equations 6 and 7.
  • Equation 16 The terms on the right side of Equation 16 can be obtained as follows.
  • K of (K / S) p can be obtained by the above equation 2 from the spectral reflectance R ( ⁇ ) of the pigment physical property determination sample (sample Ml 2 in FIG. 7) which does not contain white pigment.
  • the value of (KZS) in Equation 16 is a value for a colored pigment that does not include the effect of the base glaze.
  • (KZS) ⁇ obtained by actual measurement as described above is for the sample of (base glaze + colored pigment). The influence of the glaze is impaired.
  • the mixing ratio of the pigment of 17 samples M12 is 12%, which is a ⁇ ⁇ value, the contribution of the base glaze to (KZS) p is extremely small. Therefore, as the value of (K / S) p in Equation 16, even if the value obtained from Equation 2 from the spectral reflectance R ( ⁇ ) of sample M 12 is used, the error is negligible. is there.
  • (KZS) value of Micromax is the spectral reflectance of the white pigment and Complex both colored pigments unpigmented physical properties determined for samples (samples of Figure 7 ⁇ 1 ⁇ VI 1 1! R (;.) From the formula 2 Can be calculated according to Therefore, (KZS) ⁇ is determined in a form depending on the mixing ratio C p of the colored pigment.
  • FIG. 9 is a graph showing an example of the dependence of the absorption coefficient ⁇ ⁇ ( ⁇ ) on the mixing ratio C p .
  • the scattering coefficient S p (; J shows a similar dependence. Note that FIG. 9 shows only a graph of typical wavelengths, but in actuality, the actual wavelength range of visible light (about 400 nm ⁇ about 7 00 nm) absorption coefficient for each ⁇ 0 nm in ⁇ ⁇ ( ⁇ ) is obtained.
  • the absorption coefficient K w , (s) and the scattering coefficient S w ′ of the mixture of (base glaze + white pigment) and the absorption coefficient K p of other colored pigments and the scattering coefficient S p (; J When is obtained, in step S26 in FIG. 6, a simulation is performed on the verification sample (FIG. 8) by computer color matching.
  • Equation 18 the spectral reflectance R of the mixture is given by Equation 18 below.
  • Equation 19 the spectral reflectance ( ⁇ ) of the mixture can also be determined from Equation 18. Since this spectroscopic reflectance R ( ⁇ ) is an ideal state spectral reflectance, the spectral reflectance ( ⁇ ) that can be measured by a spectrophotometer is calculated according to the following Equation 19, which is a modification of Equation 9 (Sanderson's equation). Ask.
  • the tristimulus values X, ⁇ , ⁇ of the mixture are determined by the following equation 20.
  • S is the spectral distribution of the standard light
  • X (e), y ( ⁇ ), ⁇ ( ⁇ ) are color matching functions.
  • the tristimulus value is calculated by the above-described procedure assuming the blending ratio of the mixture, and the desired value is calculated by a successive approximation method such as the Newton-Raphson method. Find a blending ratio that matches the color of the product within a specified error.
  • the compounding ratio of each of the verification samples D1 to D4 shown in FIG. 8 is predicted by computer color matching. Since the mixing ratio of the verification sample is known, the degree of coincidence between the predicted mixing ratio and the actual mixing ratio can be easily calculated. For example, the root mean square error given by the following equation 21 is used as an index of the degree of coincidence of the mixing ratio.
  • Equation 7 coefficients in Equation 7 as cut with accurately predicting the formulation ratio of the verification samples in the first embodiment d, since the determined e, Formulation ratio C w of the A color pigment is relatively low area
  • the scattering coefficient S w 'and the absorption coefficient K w ' ( ⁇ ) of the mixture of (base glaze + white pigment) could be determined with high accuracy.
  • FIG. 10 is an explanatory diagram showing a prediction result of the mixing ratio in the first embodiment.
  • the prediction result of the first embodiment in FIG. 10 is a result obtained when Expressions 11 and 22 are used.
  • the blending ratio that minimizes the color difference was predicted using the L * a * b * color system.
  • the prediction result of the Example] has a significantly smaller root mean square error of the mixing ratio than that of the comparative example, indicating that the mixing ratio can be predicted with higher accuracy.
  • the mixing ratio of the white pigment is small as in the samples P2 and P3, the prediction accuracy of the comparative example is considerably deteriorated, whereas the prediction accuracy of the first embodiment is good.
  • the computer is used. High-precision prediction can be performed by color matching. In addition, it is only necessary to prepare a sample as shown in Figs. 4, 7 and 8 on which a mixture is applied on a normal substrate, so that it is easy to prepare a sample.
  • step S23 it is not necessary to find the coefficients by the successive approximation method, and it is possible to assume various coefficient values and select the coefficient that optimizes the simulation result in step S26. ,.
  • Equation 5 when using Equation 5 as it is, set several values of S s in the range of about 0.001 to about 0.010, and in Step S26, the value closest to the actual mixing rate is set. the compounding ratio may be so that to determine the value of S B predictable.
  • the function for expressing the scattering coefficient S is not limited to Equations 5 to 7.
  • the scattering coefficient S w ′ may be expressed by a function f (C w ) that depends on the mixing ratio C w .
  • Equation 5 can be expressed by the following Equation 23 including the constant 3l .
  • the second embodiment described below corrects the absorption coefficient Ki and the scattering coefficient S i of each component.
  • the aim is to reduce the prediction error without any change.
  • FIG. 1 is a flowchart illustrating an overall procedure of a process according to a second embodiment.
  • the mixture targeted in the second embodiment is a glaze for covering the surface of the ceramic body.
  • the base glaze (base glaze) without pigment is the object to be colored, and the glaze obtained by adding pigment to this base glaze is the mixture to be subjected to computer color matching.
  • step S31 learning of a neural network for correcting the prediction result (tristimulus value) of computer color matching is performed.
  • steps S32 to S35 the target value of computer color matching is corrected using the trained neural network to obtain an accurate prediction result.
  • FIG. 12 is an explanatory diagram showing the configuration of the neural network.
  • This dual-purpose network has a three-layer hierarchical structure composed of an input layer 10, a middle shoulder 20 and an output layer 30.
  • the human stratum I0 is composed of three neurons N11 to N13
  • the middle layer 20 is composed of five neurons N21 to N25
  • the output shoulder 30 is composed of three neurons N31 to N33. Have been.
  • Two: stimulus values X, ,, and Z are input to the three neurons N11 to N13 of the input layer 10 respectively.
  • the signal transmitted from the neuron Nij of the input layer 10 to the neuron Nk of the hidden layer 20 is obtained by multiplying each input signal by weights W and k .
  • i is a number indicating the debris of interest
  • j is a number indicating the order of neurons in the layer of interest
  • k is a number indicating the order of neurons in the next layer.
  • the signal transmitted from the first neuron N il of the input ⁇ 10 to the first neuron N 21 of the middle layer 20 is W u .
  • the first neuron N il of the input dust 10 is signal transmitted to the second neuron N 22 of pressurized et intermediate layer 2 0 is W u, 2 X.
  • ⁇ th neuron N (i-l) j in the (i ⁇ 1) th hierarchy that is, the input layer 10
  • k is the weight applied to the signal transmitted from the j-th neuron N (i-1) j in the (i-1) th hierarchy to the neuron Nij of interest.
  • t is a threshold, some other being assigned.
  • the information transfer function f (u, in Equation 24 is called a sigmoid function.
  • Equation 24 when Equation 24 is applied to the input / output relationship of the second neuron N21 of the middle layer 20, the following Equation 25 is obtained.
  • u 2 i W n , iX + W, 2jY + W 131 Z-.. (25b)
  • the input / output relationship of each neuron of the output layer 30 is also given by the above equation (24).
  • the output of the three neurons N31 to N33 in the output layer 30 is
  • the to Q 13 predicts errors ⁇ tristimulus values by computer color one matching (CCM), ⁇ , and Derutazeta.
  • Neural network learning gives a number of relationships between the inputs (X, Y, Z) to the input debris 10 and the outputs ( ⁇ , ⁇ , ⁇ ) from the output layer 30 to give the correct input / output relationships. This is the task of determining the values of the weights W and k .
  • FIG. 13 is a flowchart showing the detailed procedure of step S31 in FIG.
  • step S41 a plurality of tristimulus values (Xit. Yit, Zit) that cover the color range of the mixture to be predicted in computer color matching
  • the blending ratio of a plurality of samples having the plurality of tristimulus values is determined by computer color matching.
  • FIG. 14 is a conceptual diagram illustrating a color prediction target range PA in computer color matching and a distribution of a plurality of tristimulus values covering the prediction target range PA.
  • colors are expressed in the CIE-XYZ color system
  • the color prediction target range PA can be obtained as a three-dimensional range in the XYZ coordinate system.
  • the prediction target range PA indicates the range of colors that the mixture to be predicted can take, and is a range that can be set arbitrarily.
  • step S41 the compounding ratio of the seven types of samples M1 to M7 having these seven sets of tristimulus values (Xit, Yit, Zit) was further predicted by computer color matching.
  • Equation 26 the spectral reflectance R (s) of the mixture is given by Equation 26 below.
  • the ratio (KZS) ⁇ between the absorption coefficient and the scattering coefficient of the M mixture can be calculated from the absorption coefficient Ki (1) and the scattering coefficient S ⁇ ) of each component and the mixing ratio Ci according to Equation 1 described in the first embodiment.
  • the spectral reflectance R ( ⁇ ) of the mixture can be obtained from the above equation 26. Since this spectral reflectance R ( ⁇ ) is the spectral reflectance in an ideal state (when the thickness of the object to be colored is infinite), the spectral reflectance R '(;-) that can be measured by a spectrophotometer is expressed by the following equation. 2 7 (Sanderson's equation)
  • R ' k 1 -Kl -ki) (l -k 2 ) l ..- (27) where the coefficient, k 2 is a value that depends on the optical properties of the object to be colored (base glaze).
  • the coefficients, k 2 can be determined from the refractive index n of the object to be colored according to the following equation 28.
  • the base glaze used in the second example had a refractive index n of about 1.4.
  • the spectral reflectance R ') is determined by Equation 27
  • the tristimulus values X, Y, and Z of the mixture are determined by Equation 29 below.
  • S ( ⁇ ) is the spectral distribution of the standard light
  • X (/), y ( ⁇ ), and ⁇ ( ⁇ ) are color matching functions.
  • the tristimulus values X, ⁇ , ⁇ Is calculated. Since the tristimulus values represent the color of the mixture, the color of any mixture can be predicted.
  • the tristimulus value is calculated by the above-described procedure assuming the blending ratio of the mixture, and the desired value is calculated by a successive approximation method such as the Newton-Raphson method.
  • Formula 1 is determined so that the color matches the color within a specified error.
  • step S42 of FIG. 13 a plurality of samples having the compounding ratio C i predicted in step S4 ⁇ are created.
  • seven samples M1 to M7 corresponding to the seven sets of tristimulus values (Xit, Yit, Zit) in FIG. 14 were created.
  • step S43 the spectral reflectance of each sample Mi is measured with a spectrophotometer, and its tristimulus values (Xim Yim, Zin) are obtained according to the above equation (29).
  • step S44 for each sample Mi, the difference between the measured value (Xim, Yim, Zim) obtained in step S43 and the target value (Xit, Yit, Zit) determined in step S41 is calculated.
  • FIG. 15 is a conceptual diagram illustrating prediction errors of seven samples M1 to M7 obtained in the second example.
  • FIG. 16 is a description “ ⁇ 1” showing the target value of the tristimulus value (the value determined in step S41) of each sample and the prediction error ⁇ ( ⁇ , ⁇ , ⁇ ).
  • the predicted values (Xic, Yic, Zic) of the tristimulus ⁇ of each sample were obtained from the mixture ratio C i obtained in step S41 by computer color matching, and the measured values (Xim, Yim, Zim) were obtained.
  • the difference (Xim—Xic, Yim-Yic, Zim—Zic) between the prediction error and the prediction ⁇ (Xic, Yic, Zic) may be defined as the prediction error AMi.
  • the mixing ratio C i is determined so that the difference between (Xit, Yit, Zit) and the predicted value (Xic, Yic, Zic) is equal to or less than a predetermined allowable error
  • the predicted values (Xic, Yic, Zic) have substantially the same value. Therefore, even if the difference (X im-Xic, Yim—Yic, Zim—Zic) between the measured value (Xim, Yim, Zim) and the prediction ⁇ (Xic, Yic, Zic) is defined as the prediction error ⁇ , Measured value
  • step S45 in FIG. 13 learning of the neural network is performed using the target values (Xit, Yit, Zit) and the prediction error ⁇ ⁇ ⁇ ⁇ ⁇ of the tristimulus values of each sample Mi shown in FIG. Determine the weights Wij, k at.
  • a learning method of the neural network for example, a back error propagation learning method is used.
  • FIG. 17 is an explanatory diagram showing the results of the verification of the neural network learning in the second embodiment.
  • another set of tristimulus values included in the prediction target range PA shown in Figs. 14 and 5 is set as the target value for computer color matching.
  • An eighth sample M8 having a standard value was prepared. Then, the sagittal depression of this sample M8 was actually measured.
  • CCM target value means the target value used for the computer color matching of the eighth sample M8.
  • error (true planting)” is the difference between the target value and the measured value of the tristimulus value.
  • Neuro prediction error is the prediction error obtained when the CCM target value is input to a trained neural network (Fig. 12). From the results in Fig. 7, it can be seen that the trained bi-ural network can accurately predict the error of tristimulus values.
  • steps S32 to S35 in FIG. 11 are executed to predict the mixing ratio of a color sample whose mixing ratio is unknown.
  • step S32 the spectral reflectance of the color sample whose mixing ratio is unknown is measured, and its tristimulus value (Xs, Ys, Zs) is obtained.
  • step S33 the tristimulus values (Xs, Ys, Zs) of the color samples are input to the neural network shown in Fig. 12, and the prediction error ⁇ (AXs, AYs, ⁇ s) is obtained. .
  • step S34 the tristimulus values (Xs, Ys, Zs) are corrected by the prediction error, and the target of the tristimulus values ⁇ (Xs— ⁇ s, Y s - ⁇ s, Zs—AZs).
  • step S35 computer color matching is performed using the corrected target value to predict the mixing ratio of the color sample.
  • FIG. 18 is an explanatory diagram showing the results of an experiment performed to verify the prediction accuracy of computer color matching in the second embodiment.
  • we aimed to verify the prediction accuracy so we measured the tristimulus values for a color sample with a known blending ratio, and calculated the blending ratio that realized the measured values (Xs, Ys, Zs) using a computer. Predicted by color matching. From the results in Fig. 17, it can be seen that the true value of the blending ratio and the predicted value by computer color matching are in very good agreement.
  • the learned neural network shown in FIG. 12 it is possible to accurately predict the tristimulus values of the mixture with a known mixing ratio. That is, tristimulus values are obtained from the mixture ratio C i of the mixture to be predicted using the above formulas, 26 to 29, and the tristimulus values are input to the neural network shown in FIG. Ask. Then, if the tristimulus value obtained by Equation 29 is corrected by the prediction error, it will be extremely close to the actual value. Very close: Stimulus value is obtained.
  • the theory of color mixing may not be established due to firing conditions, chemical reaction of raw materials during melting, and the like.
  • the method of this embodiment in which the neural network learns the relationship between the tristimulus value of the mixture and the prediction error due to CCM is effective.
  • the target values of the tristimulus values in the computer color matching were input to the neural network.
  • the human power of the neural network the measured tristimulus values (Xim, Yim, Zim) may be used, or predicted values (Xic, Yic-, Zic) by computer color matching may be used.
  • the neural network may be made to learn the relationship between the tristimulus values (coordinate values in the XYZ color system) of a plurality of samples and their prediction errors.
  • the tristimulus values are corrected using a neural network, but may be corrected using other error correction methods such as regression analysis or neuro-fuzzy technology. is there.
  • Industrial porcelain such as sanitary ware
  • the glaze corresponds to the mixture in computer color matching.
  • the designer first determines the color by coloring the paper or the like or selecting a sample having a desired color from pre-baked samples. Then, the spectral reflectance of the color is measured with a spectrophotometer, and the mixing ratio of a coloring agent such as a pigment or a dye is predicted from the measured value of the spectral reflectance using computer matching.
  • a coloring agent such as a pigment or a dye
  • Figure 9 is an X-Y chromaticity diagram showing the variation between the designed colors and the colors of the pottery actually manufactured.
  • the color indicated by the double circle is the designed color (standard color) L0
  • the small white circle is the color distribution of the ceramics actually manufactured. Pottery with a color that is significantly different from the standard color L0 must be recognized as defective. Therefore, the shade limit colors L1 and L2, which have the colors indicated by the black circles in Fig. 19, are set to indicate the shade limits of non-defective products.
  • the chromaticity coordinate values of the standard color L0 are determined by the designer, the chromaticity coordinate values of the standard color L0 are taken into consideration by considering the manufacturing error and the like, and the two 'shade limit colors L1, L2' Determine the chromaticity coordinate value of.
  • a pottery having the standard color L0 is created as a standard color sample, and two potteries having shading limit colors L I and L2 are created as shading limit samples. Then, in the inspection process, the color of the standard color sample and the color limit sample is compared with the color of the manufactured ceramic, and only the ceramics in the color range of the color limit sample are regarded as good products.
  • the third embodiment described below aims at accurately estimating the blending ratio of a mixture having a desired color.
  • FIG. 20 is a flowchart showing the overall procedure of the process in the third embodiment.
  • the mixture targeted in the third embodiment is a glaze for covering the surface of a ceramic body.
  • the base glaze without pigment (base glaze) is the object to be colored, and the glaze obtained by adding pigment to this base glaze is the mixture to be subjected to computer color matching.
  • step S51 first, a standard color sample is created. Therefore, the mixing ratio of the colorant of the standard color sample is known.
  • step S52 the spectral reflectance R 'of the created standard color sample is measured with a spectrophotometer, and the tristimulus value X is calculated from the spectral reflectance R' according to the following equation (30). , ⁇ . , ⁇ Calculate () .
  • S ( ⁇ ) is the standard-light spectral distribution
  • X ( ⁇ ), y) and ⁇ ( ⁇ (with a bar in the formula) are color matching functions.
  • step S53 the measured value X of the tristimulus value of the standard color obtained in step S52. , ⁇ . , Z.
  • the designer Set the tristimulus values for the shade limit colors LI, L2 of.
  • a blending ratio for realizing the first shade limit L ⁇ is predicted.
  • the blending ratio can be accurately predicted by the same processing.
  • step S54 the standard color sample is obtained from the mixing ratio Ci (i indicates the number of the colorant), the scattering coefficient Si, and the absorption coefficient Ki of each colorant of the standard color sample created in step S51.
  • the calculated values X M , Y M , and Z N of the tristimulus values are calculated by the following procedure.
  • the spectral reflectance R ( ⁇ ) of the mixture is given by the following Equation 31.
  • the ratio (KZS) ⁇ between the absorption coefficient and the scattering coefficient of the mixture can be calculated from the absorption coefficient ( ⁇ ), the scattering coefficient Si (), and the blending ratio Ci of each colorant according to Equation 1 described in the first embodiment.
  • the spectral reflectance R of the mixture (lambda) from the ratio (KZS) Micromax can be determined according to equation 3 1.
  • the thickness of the spectral reflectance R (/) is an ideal state (the colorings infinite )
  • the spectral reflectance R '( ⁇ ) that can be measured with a spectrophotometer is calculated according to the following equation 32 (Sanderson's equation).
  • R ' k 1 + (lk 1 ) (lk 2 ) 1 ... (32)
  • k 2 is a value that depends on the optical properties of the object to be colored (base glaze).
  • the refractive index n of the base glaze is, for example, about 1.4.
  • the tristimulus values X-, Y M and Z H can be calculated according to the same formula as the above-described formula 30.
  • the absorption coefficient and the scattering coefficient S i of each colorant include an error, and the equations 32 to 33 are empirical rather than theoretical, so the tristimulus values obtained in step S 54 are calculated.
  • the values X M , Y M , and Z M have errors.
  • This calculation error is the measured value X obtained in step S52. , Y. , ⁇ . Is the difference from
  • the prediction of the blending ratio by computer color matching also includes the step of calculating the tristimulus values of the mixture according to the formulas 1, 2, 30 to 33. Therefore, when predicting the mixing ratio of the shading limit sample by computer color matching, if the calculation error of the tristimulus value in the standard color sample is considered, the prediction accuracy can be improved.
  • step S55 of FIG. 20 an actual measurement ⁇ X of the tristimulus value of the standard color obtained in step S52. , Upsilon 0, and Zeta 0, Step S 5 4 out was calculated values ⁇ ⁇ , ⁇ ⁇ , difference ⁇ with ⁇ ⁇ , ⁇ , seek ⁇ following Equation 3 4.
  • step S56 the tristimulus values X ⁇ , ⁇ .,., ⁇ ⁇ set in step S53 are corrected by the differences ⁇ , ⁇ , ⁇ of the above ti. target value X c tristimulus values in computer color matching, Y c, Ru seek Z c. That is, the target values X c , Y c , and Z c of the tristimulus values for the lysis limit sample are given by the following Expression 35.
  • step S57 the mixing ratio of the density limit sample is predicted by computer color matching.
  • the mixing ratio is determined so that the target values Xc , ⁇ c, and Zc given by Expression 35 are obtained.
  • FIG. 21 is a flowchart showing the detailed procedure of the computer color matching in step S57. The procedure in Fig. 21 applies computer color matching using the Newton-Raphson method.
  • step S6 the change in tristimulus value when the mixing ratio Ci of each colorant (pigment) is slightly changed is calculated in the following procedure.
  • the differences AX Ci , AY ci , and AZ Ci from the calculated values ⁇ , ⁇ ,, ⁇ ⁇ of the tristimulus values of the standard color obtained in step S 52 are calculated according to the following equation 36. .
  • Equation 36 the change rate of the tristimulus value when only the mixing ratio Ci of each colorant is slightly changed is given by the following Equation 37. )
  • Equation 38 assumes that there are four types of colorants. Note that a fixed value is substituted for the total value ⁇ CV of the change amount of the mixing ratio of each colorant. For example, if the sum ⁇ Ci of the mixing ratios Ci of the four colorants is kept constant,? Is 0.
  • step S63 using him obtained in step S62 , the mixing ratio C iT of each colorant of the shading limit sample is calculated according to the following Expression 39 .
  • step S 64 from the blending ratio C iT obtained in step S 63, the number mentioned above Equation 1, 2, calculated values of tristimulus values according to. 30 to 3 3, Upsilon chi, seek Z 2.
  • scan Tetsupu S 65 thus resulting tristimulus values, Upsilon:, Z: the target value X c tristimulus values obtained in scan Tetsupu S 5 6 of FIG. 20, Y c, the color difference between Z c It is determined whether or not the predetermined tolerance is less than or equal to ⁇ 5.
  • the color difference is given by the following equation 40 using the Lab color system.
  • step S65 if the color difference ⁇ ⁇ ⁇ is smaller than the allowable error S, The value of the mixing ratio CiT obtained in step S63 is adopted as the predicted value, and the computer color matching is terminated.
  • the color difference E is equal to or larger than the allowable error ⁇
  • the calculated values ⁇ 1 , ⁇ 1 , ⁇ , of the tristimulus values obtained in step S 64 are replaced with ⁇ ⁇ , ⁇ ,, ⁇ in step S 66.
  • steps S62 to S65 is repeated.
  • FIG. 22 is a table showing tristimulus values of a standard color sample and tristimulus values of a shading limit sample used in the third embodiment of the present invention.
  • Samples T l, ⁇ 2, ⁇ 3 are standard color samples having different colors.
  • Fig. 22 shows the tristimulus values X actually measured in step S52 in Fig. 20 for each standard color sample. , Upsilon 0, and Zeta ,,, Step S 5 4 tristimulus values chi Micromax calculated in, Upsilon Micromax, are shown and Zeta Micromax is also shading boundary samples set in Sutetsu flop S 5 3
  • the set values of the tristimulus values ⁇ ⁇ , ⁇ ⁇ , and ⁇ ⁇ are also shown.
  • the target values X c , Y c , .Z c of the tristimulus values of the gray scale limit sample in the computer color matching are calculated from these values according to the above-mentioned equations 34
  • FIG. 23 shows the prediction results of the tristimulus values of the density limit sample in the third embodiment, the prediction results of the comparative example, and the actual blending ratio.
  • the blending ratio of the comparative example is a predicted value obtained when the set values ⁇ ⁇ , ⁇ ⁇ , ⁇ ⁇ of the tristimulus values of the gray scale limit sample are directly used as the target values of the computer color matching.
  • compounding ratio of the third embodiment is the predicted value obtained using the target value X c were corrected in the calculation error of the standard color swatch, Y c, the Z c.
  • a gray scale limit sample was actually manufactured using the mixing ratio shown on the right end of FIG. 23, and the measured tristimulus values were compared with those in FIG. Are used as the set values ⁇ ⁇ , ⁇ ⁇ , ⁇ ⁇ shown in Fig.
  • the third example was able to predict the mixing ratio with higher accuracy than the comparative example.
  • the prediction accuracy of the third embodiment is higher than that of the comparative example in all cases.
  • the calculation error of tristimulation was calculated for a standard color sample having a tristimulus value close to the filtering limit sample, and this was used to calculate the tristimulus value of the shading limit sample. The target value has been corrected. As a result, it is possible to improve the prediction accuracy when predicting the mixing ratio of the shading limit sample.
  • the standard-color sample is used as a sample having a tristimulus value close to the density limit sample (near-color sample), and the calculation error of the tristimulus value for the standard color sample is used.
  • the target value of the tristimulus value of the light-limit sample was corrected.
  • a sample having a color close to the shading limit sample may be selected as a nearby color sample.
  • This database should preferably include at least the formula C i of each colorant and the measured tristimulus value (or the reflectance R ( ⁇ )).
  • CCM computer color matching
  • the object to be colored is a fiber, and if the color of the dyed (dyed) fiber is not the desired one, the dyed fiber is additionally dyed to change the color to the desired one. I do.
  • the color is determined by the mixing ratio (mixing ratio) of the pigment mixed with the glaze, and the color is obtained through firing of the glaze (hereinafter referred to as glaze) in which the pigment is mixed.
  • glaze the glaze in which the pigment is mixed.
  • the above-mentioned conventional CCM cannot be applied to an object to be colored that cannot be additionally dyed (colored), such as pottery and tiles.
  • the colorant (dye for fiber, glaze for pottery and tile) is used for fiber, pottery, etc.
  • the mixing ratio of the colorant is maintained, but if the mixing process fluctuates, for example, if there is a deviation in the temperature, the timing of the mixing, or the like, the color obtained from the colorant previously mixed or the color of the color sample is changed. There is a case that cannot be redone. Especially in ceramics and tiles, the reliability of color reproduction is somewhat lacking due to the use of natural pigments.
  • the color can be corrected by additional dyeing using the above-mentioned CCM.
  • the colorant that has been prepared must be used in combination with the already prepared colorant. It is complicated. In order to avoid this complication, a colorant that can obtain a color sample color may be newly prepared and dyed with only this colorant. However, the already prepared colorant is discarded as unnecessary or discarded. It is necessary to add a new colorant to the prepared colorant and re-formulate. Also, for pottery and tiles, it is not possible to add additional color, so the prepared colorant must be discarded or remixed. But However, disposing of the prepared colorant is useless, while re-formulation of the colorant is also cumbersome because the colorant is gradually added based on the intuition and experience of engineers.
  • the fourth embodiment described below aims at simplifying re-mixing while effectively utilizing the mixed colorant.
  • FIG. 24 is a block diagram illustrating an apparatus for performing the computer color matching method according to the embodiment.
  • This device can also be used as a device that implements the first to third embodiments described above.
  • the arithmetic unit 40 is a general-purpose computer, and realizes each step and each means of computer color matching according to the present invention by causing a CPU (not shown) to execute a software program.
  • FIG. 25 is a flowchart showing the entire procedure of the process in the fourth embodiment.
  • the composition targeted in the fourth embodiment is a glaze for covering the surface of a ceramic body.
  • the base glaze without pigment (base glaze) is the material to be colored, and the glaze with the pigment added to this base glaze is the target for computer color matching.
  • the color system is the XYZ color system, it goes without saying that other color systems, for example, the L * a * b * color system may be adopted.
  • an arithmetic unit 40 for executing a rare operation which will be described later, relating to the CCM
  • an input device 42 such as a keyboard and a mouse for inputting data
  • a display device 44 for displaying the results of pass / fail judgment described later
  • a storage device 46 for storing the results of the pass / fail judgment and various arithmetic expressions, etc., and acquiring a spectral reflectance as data necessary for CCM.
  • a spectrophotometer 48 is used. Then, in the arithmetic unit 40, the following blending process is performed.
  • step S71 a glaze (sample glaze) exhibiting a target color is prepared. Since the H standard color in this case is the color presented by the prepared glaze, the blending ratio of the pigment in this glaze (sample glaze) is known.
  • step S72 the ⁇ standard color sample glaze is measured with a spectrophotometer 48, and tristimulus values (actually measured values) ⁇ ⁇ , , ⁇ , ⁇ ⁇ which are color evaluation values in the XYZ color system Ask for.
  • the tristimulus value is calculated from the spectral reflectance R ′ (-) of the target color sample glaze obtained by colorimetry with a spectrophotometer 48 according to the following equation 42. Is done.
  • the calculated tristimulus value is displayed on the display device 44 together with the target sample color, and is stored in the storage device 46 for use in processing described later.
  • the calculation results of the tristimulus values and the like described below are stored in the storage device 46 each time.
  • each pigment is added at a blending ratio assumed to exhibit a color similar to the target color exhibited by the sample glaze, and the second trial glaze is blended.
  • the color is measured in the same manner as in S72.
  • the tristimulus values (actually measured values),, and Z for the first prototype glaze are obtained.
  • the pigments are blended at the above blending ratios by a technician, but the values are arbitrary values that are known and need to be repeatedly blended by trial and error as in the past. No special experience or intuition is required to determine the mix ratio.
  • the tristimulus value and the pigment preparation amount (preparation ratio) for the first prototype glaze are also displayed on the display device 44 and stored in the storage device 46.
  • the amount of the mixture at this time is input from the input device 42.
  • the color difference ⁇ * (JIS Z 8 "730) between the color of the target color sample glaze created in step S71 and the color of the first prototype glaze in step S73 is set within a predetermined range.
  • the allowable value of the color difference ⁇ * is determined by the difference between the color of the target color sample and the color of the first prototype glaze. Value that cannot be distinguished by The value is set in advance from the manpower device 42.
  • the allowable value of the color difference ⁇ * can be set to a value other than 0.3 to 0.5.
  • step S74 If it is judged that the color difference ⁇ * is within 0.3 to 0.5 in step S74, the color of the target color sample can be re-used with the glaze of the first prototype in step S73. It is determined that no further compounding processing is necessary, and all the processing ends. In other words, the new glaze blended at the blending ratio at the time of the first trial production has almost the same color as the target color sample glaze.
  • step S74 if the color difference ⁇ * does not fall within the range of 0.3 to 0.5 in step S74, the rejection is determined, and in step S75, the target color obtained in step S72 is determined.
  • the difference between the tristimulus values according to the following equation 43 is ⁇ , ⁇ , ⁇ .
  • the differences ⁇ , ⁇ , ⁇ of the tristimulus values reflect the color difference ⁇ * between the color presented by the target swatch and the color exhibited by the first prototype glaze.
  • step S76 the rate of change in tristimulus value ( ⁇ coefficient) when a small amount of each pigment (mixing ratio is known) is added to each pigment (mixing ratio is known) at the time of the first trial glaze blending performed in step S73 Is calculated using the CCM method according to the following procedure. It should be noted that a slight increase in the amount of each additional pigment to be prepared is also input from the input device 42.
  • the tristimulus value of the color exhibited by the first prototype glaze was calculated for each pigment at the time of the first prototype. Calculate using the CCM method from the known blending ratio.
  • the tristimulus values (calculated values) X 1 / E , Y 1 / E , and Z 1 / E in this case are represented by the following Equation 44.
  • the absorption coefficient ⁇ ?.) And the scattering coefficient Si (.) Of the coloring matter and the coloring agent are expressed by the following Duncan's equation represented by the following equation 45 and Kubelka-munk represented by the equation 46. -Munk), the spectral reflectance RU) of any formulation can be determined by CCM based on these equations.
  • the ratio between the absorption coefficient and the scattering coefficient (KZS) of the preparation is represented by the absorption coefficient Ki (A), the scattering coefficient Si (;.) And the mixing ratio Ci of each pigment, as shown in Equation 47. Is defined by Therefore, the spectral reflectance R ( ⁇ ) can be calculated and calculated from this ratio (KZS).
  • K w and S w are the absorption coefficient and scattering coefficient of the white component (white pigment), and C w is the blending ratio of the white component.
  • step S92 for each pigment (pigments 1, 2, and 3), the tristimulus value (calculation) for the color of the glaze, in which the proportion of the first prototype glaze was separately increased by a small amount, was calculated.
  • Value) X1 / 1 / E , Y1 / 1 / E , Z1 / 1 / E were calculated from the known blending ratio of each pigment at the time of the first trial production and the known blending ratio of the increased pigment. Calculate using the method described above. To explain in more detail, first, add a small amount of pigment 1 (0.1 * C,) to the first trial glaze, and add other pigments 23 and white pigment to the glaze with the same blending ratio as the first trial glaze. The tristimulus value for the color presented by is calculated. Even in this case, the above Equation 44 Equation 47 is used, and in Equation 47, +0.1
  • Pigment 2 is added in only a small amount (0.1 * C 2 ), and the other pigments are tristimulus values X 1/2 / E Y 1/2 / E Z for the color of the glaze of the same mixing ratio. 1/2 / E , and pigment 3 with only a small amount (0.) * C 3 ), and the other pigments have the same stimulus tristimulus value for the color of the glaze X 1/3 / E Calculate Y 1/3 / E Z 1/3 / E.
  • the tristimulus value X 1 / w / E Y 1 / w / E z 1 / w / E is similarly calculated for the white pigment.
  • Pigment 1 Minor increase formulation: d6fcvJ
  • step S77 it is necessary to correct the differences ⁇ ⁇ , ⁇ , ⁇ (actual measurement values) of the tristimulus values obtained in step S75 through colorimetry of the target color sample glaze and the first prototype glaze.
  • the required additional compounding amount of each pigment is calculated using the following Equations 49 and 50.
  • Equation 4-9 the difference between the tristimulus values ⁇ , ⁇ , provided for correcting ⁇ (difference measured value), additional bulking formulation amount of each pigment (ACi, AC 2, AC 3 , AC W) Is a variable.
  • Equation 50 is a cost function when each pigment is varied, and the amount of addition of each pigment ( ⁇ ( ⁇ , AC 2 , m C 3 , AC W ) is also a variable.
  • equation 4 9, 5 0 these variables by solving the four above variables, i.e. additional increase formulation amount of each pigment ( ⁇ ( ⁇ , AC 2, AC 3 , AC W) is determined.
  • equation 5 0 is an expense required to a unit variation amount of the pigment i.
  • step S 7 in this case, in step S 7 7, ⁇ satisfies ⁇ 0, AC W satisfies AC W ⁇ 0, and, in equation 5 0 the AC W a cost function F to delta ⁇ sequence that minimizes represented solved using linear programming. Accordingly, and never AC W is calculated as a negative value.
  • the blending ratio of each pigment is calculated by taking into account the determined amount of added bulk compounding, and the physical property values of the additionally blended glaze are calculated and updated. I do. More specifically, the spectral reflectance R (;-) is obtained from the formulas 45 to 47 using the blending ratio, the absorption coefficient and the scattering coefficient of each pigment, and the like. Then, in the subsequent step S79, the calculated spectral reflectance R ( ⁇ ) is substituted into Equation 44, and the tristimulus value (calculated value) for the color of the glaze that is obtained by adding and increasing each face bridge is added.
  • step S80 the tristimulus values X END / E , Y END / E , Z END / E. And the tristimulus values ⁇ ⁇ , ⁇ ⁇ ⁇ ⁇ for the target swatches are used.
  • JIS Z 8730 the pass / fail judgment is again made as to whether or not the color difference ⁇ ⁇ * falls within a predetermined range. If it is judged that the color difference ⁇ * is within 0.30.5 in this step S80, if the glaze prepared with the correction amount obtained in the previous steps reproduces the color of the target color sample You can do it. Therefore, in this case, the process is terminated assuming that the blending process is completed.
  • the blending ratio that takes into account the correction amount (additional blending amount) obtained for each pigment is used as the final blending ratio when the glaze that has been blended is reblended. More specifically, the blended glaze whose color has been modulated due to variations in the blending process, etc., corresponds to the first prototype glaze in the fourth embodiment because the blending ratio is known, If each pigment is added to the blended glaze whose color has been modulated and added at the blending ratio according to the correction amount described above, the glaze exhibiting the target color can be blended again.
  • the tristimulus values for differential coefficient calculation and correction amount calculation are set to the previous values. And change the slight increase of each pigment. More specifically, the tristimulus value X BND / E Y BND / E , which is obtained in step S79 , is shifted to the tristimulus values, ⁇ ,, and Z X in Expression 43. From the tristimulus values (X / E , YEND / E , Z / E ) after this shift, the difference (mmX ⁇ , ⁇ ) between the tristimulus and the target color sample is newly obtained in step S75. Is calculated.
  • step S76 adds the microaddition amount considered in step S76 to the previous formulation amount (in this case, pigment 1 is Ci + 0.1 ⁇ ). Change to a small increase (( ⁇ + 0.1) * 0.1) multiplied by 1. This allows each pigment to be 0. 0 each less than if it were rejected in step S80. Since the amount is increased by a small amount, in step S76, the amount of each pigment is ⁇ tft, based on the proportion of the fine ftif that has been reduced. It is required according to.
  • this new tristimulus value due to the slight increase of each pigment is used in place of the tristimulus values X 1/1 / E , ⁇ 1/1 / ⁇ , Z 1/1 / E, etc. in Equation 48 .
  • the derivative is calculated again. After that, as described above, the amount of additional compounding (correction amount) of each pigment is determined, and the above-described processing is repeated until a pass is determined in step S80.
  • the glaze target color sample glaze
  • the glaze that has existed up to that point has a blending ratio such that the color presents a color that is somewhat similar.
  • Ask the amount of negative compounding that means the removal of pigment is not determined by CCM. Therefore, according to the computer color matching method of the fourth embodiment, it is not necessary to dispose of the glaze (colorant) in which the pigment has been prepared, so that the existing glaze can be effectively used.
  • the process involving the engineer is the one-time trial production of the glaze in step S73. At this time, since the engineer's many years of intuition and experience are not required, the glaze is used. Can be simplified.
  • the correction amount (AC ⁇ and AC W ) of each pigment is calculated by Equation 50.
  • the cost was calculated by the linear programming method using the expressed cost function F so that the cost required to correct each pigment was minimized. For this reason, according to the computer color matching method of the fourth embodiment, in addition to simplifying the above-mentioned effective use and re-mixing of the existing sleeves, the cost can be reduced.
  • Figure 27 shows tristimulus values (color values) for the target color sample glaze obtained in step S72. And the tristimulus values of the first trial glaze obtained in step S73 and the blending ratio (mixing ratio) of each pigment in the target color sample glaze and the first trial glaze.
  • ⁇ ⁇ * in the figure is a color difference between the target color sample glaze and the first prototype glaze, and a pass / fail judgment is made in step S74 based on this value.
  • Figure 28 shows the rate of change (differential coefficient) of tristimulus values when each pigment was added to the first prototype glaze shown in Figure 27 in a very small amount. It is calculated from At this time, the calculated tristimulus values X 1/1 / E , Y 1/1 / E , and Z 1/1 / E are used to calculate the differential coefficient when the amount of pigment 1 (red pigment) is slightly increased.
  • pigment 2 yellow pigment
  • tristimulus values X 1/2 / E, Y 1/2 / E, Z 1/2 / E power the pigment 3 (blue pigment) the tristimulus values X 1/3 / E, Y 3 / E , z 1/3 / E force;
  • tristimulus values X 1 / w / E , Y 1 / H / E , ⁇ 1 ⁇ / ⁇ are used.
  • FIG. 29 shows the result of the computer color matching method according to the fourth embodiment.
  • the glaze prepared at the final blending ratio determined for each pigment and the target color sample glaze are shown in FIG.
  • the stimulus values are shown in comparison.
  • the color difference ⁇ * for both glazes was 0.47, indicating that the pass was judged in step S80.
  • Equation 51 a utility function F represented by the following Equation 51 may be used.
  • the minimum value of the additional additional compounding amount of each pigment is defined in advance as the minimum additional additional compounding amount ⁇ C step-.
  • each pigment is additionally increased by an integral multiple of this ⁇ C step.
  • Each pigment is additionally added in this way, and the additional amount when the color difference ⁇ ⁇ * from the target color sample glaze becomes the smallest is used as a correction amount for calculating the differential coefficient, and this correction amount is taken into account.
  • the blending ratio determined may be used as the final blending ratio when re-mixing the blended glaze.
  • the minimum additional compounding amount AC The step is defined as the minimum increment of the minimum unit in which the color changes slightly when each pigment is added.
  • the maximum allowable amount A Craax of the additional amount of each pigment may be determined in advance, and the total amount of the additional amount of each pigment may be defined by this A Cmax.
  • the maximum allowable amount A Cmax is defined as an additional increase in the color that the color changes when the pigment is added, and it is considered that the original color cannot be returned with the additional amount of the other pigment.
  • the present invention is not limited to the linear programming using the above-mentioned function F, and other methods can be adopted.
  • the pass / fail judgment is made based on the color difference ⁇ *
  • a certain width is allowed for the pass / fail judgment. Therefore, a so-called fuzzy linear programming method that introduces a certain degree of ambiguity in this pass / fail decision can be adopted.
  • the result of the computer color matching method using the method of Fujii «Keisei ⁇ is shown in Figure 30 below. As shown in Fig. 30, the glaze blended at the final blending ratio determined for each pigment and the target color sample glaze have a color difference ⁇ * of 0.20, and both colors are well It turns out that they match.
  • the glaze for coloring pottery and tiles has been described as an example, but it is needless to say that the present invention can be applied to a dye for dyeing fibers.
  • a computer color matching method and apparatus include a method for predicting a mixing ratio of a colorant to be mixed with a glaze for coloring pottery tiles and a mixture thereof.

Landscapes

  • Physics & Mathematics (AREA)
  • Spectroscopy & Molecular Physics (AREA)
  • General Physics & Mathematics (AREA)
  • Spectrometry And Color Measurement (AREA)
  • Investigating Or Analysing Materials By Optical Means (AREA)
  • Coloring (AREA)

Abstract

A scattering coefficient (SW') for a mixture of an object of coloring and a white pigment is expressed by a function of the white pigment content (CW). Computer color matching is executed for a verification sample, and a value which gives high prediction accuracy is determined as coefficients (a, b, c and d) (or SE) contained in the function representing the scattering coefficient (SW'). An absorption coefficient of (the object + white pigment), absorption coefficients of other coloring pigments and scattering coefficients are determined by using the scattering coefficient (SW') given by the function so determined. Computer color matching is executed by using physical constants of each component obtained in this way, and the color and the matching proportion of the mixture are predicted. In computer color matching, a prediction error can be reduced by utilizing a neural network.

Description

明細害 コンピュータカラ一マッチング方法および装置  Specification damage Computer color matching method and apparatus
[技術分野] [Technical field]
本発明は、 コンピュータカラーマッチングによつて着色剤の調合割合の予測ま たは混合物の色予測を行なう方法およびそのための装置に関するものである。  The present invention relates to a method and apparatus for predicting the mixing ratio of a colorant or the color of a mixture by computer color matching.
[背景技術] [Background technology]
顔料や染料などの着色剤を被着色物に混合した混合物の色を予測するために、 いわゆるコンピュータカラ一マツチングが利用されている。 コンピュータカラー マッチングでは、 被着色物と着色剤の吸収係数 Ki (λ ) と散乱係数 ) と を用い、 ダンカン (Duncan) の式 (数式 1 ) と、 クベルカ一ムンク ( Kubelka-Munk) の混色理論による式 (数式 2) に基づいて、 住意の混合物の 分光反射率 R (/) を求めることができる。  In order to predict the color of a mixture in which a coloring agent such as a pigment or a dye is mixed with a material to be colored, so-called computer color matching is used. In computer color matching, the absorption coefficient Ki (λ) and scattering coefficient) of the coloring matter and the colorant are used, and the Duncan equation (Equation 1) and the Kubelka-Munk color mixing theory are used. Based on the equation (Equation 2), the spectral reflectance R (/) of the mixture of resident can be obtained.
(la) (la)
i=]  i =]
S\i: j j .-.(lb) S \ i: j j .-. (Lb)
i=l  i = l
(s)M~~ R~ *··(2) ここで、 KM , SM は混合物の吸収係数と散乱係数、 Ki , S± は i番目の成 分の吸収係数と散乱係数、 Ci は i番目の成分の調合率である。 但し、 この明細 書の数式においては、 波長 Iに依存していることを示す 「 ( ) 」 は省略されて いる。 混合物の成分は、 被着色物と着色剤である。 (s) M ~~ R ~ * (2) where K M and S M are the absorption and scattering coefficients of the mixture, Ki and S ± are the absorption and scattering coefficients of the i-th component, and Ci is This is the mixing ratio of the i-th component. However, in the mathematical expressions in this specification, “()” indicating that the wavelength depends on the wavelength I is omitted. The components of the mixture are an object to be colored and a colorant.
混合物の分光反射率 R (λ ) が解れば、 その混合物の三刺激値 X, Υ, Ζが計 算できるので、 混合物の色を算出することができる。 Knowing the spectral reflectance R (λ) of a mixture, the tristimulus values X, Υ, and の of the mixture are calculated. Therefore, the color of the mixture can be calculated.
コンピュータカラ一マッチングを行なう場合には、 被着色物と種々の着色剤の 吸収係数 (λ ) と散乱係数 S± (λ ) を予め求めておく必要がある。 ところカ^ 吸収係数 Ki (; J と散乱係数 Si (λ ) を求めるのは必ずしも容易ではない場合 が多い。 When computer color matching is performed, the absorption coefficient (λ) and scattering coefficient S ± (λ) of the material to be colored and various colorants must be determined in advance. However, it is often not always easy to obtain the absorption coefficient Ki (; J) and the scattering coefficient Si (λ).
被着色物と種々の着色剤の吸収係数 Ki (;. ) と散乱係数 Si ( . ) を求める方 法としては、 絶対法と相対法がある。 絶対法は、 各物質の吸収係数 Κ± (λ ) と 散乱係数 (;.) の絶対値を求める方法である。 一方、 相対法は、 基準となる 顔料 (通常は白色顔料) の散乱係数 sw を 1と仮定して、 各物質の吸収係数There are an absolute method and a relative method for obtaining the absorption coefficient Ki (;.) And the scattering coefficient Si (.) Of the material to be colored and various colorants. The absolute method is a method for calculating the absolute values of the absorption coefficient Κ ± (λ) and the scattering coefficient (;.) Of each substance. On the other hand, the relative method, a primary pigment (usually white pigment) assuming scattering coefficient s w 1 and the absorption coefficient of each substance
(λ ) と散乱係数 Si (A) の相対値を求める方法である。 数式 2に示すように、 混合物の分光反射率 R ) は混合物の吸収係数 KM と散乱係数 SM の比で与え られる。 従って、 各物質の吸収係数 Ki (; .) と散乱係数 Si (λ) の絶対値が不 明でも、 各物質の吸収係数 (λ ) と散乱係数 Si (λ ) の相対値が解れば、 数 式 1および 2から正しい分光反射率 R (Λ) を求めることができる。 絶対法に用 いるサンプルの作成はかなり困難であり、 煩雑な作業を要するので、 相対法が用 いられるのが普通である。 This is a method of calculating the relative value of (λ) and the scattering coefficient Si (A). As shown in Equation 2, the spectral reflectance R) of the mixture is given by the ratio of the absorption coefficient K M and the scattering coefficient S M of the mixture. Therefore, even if the absolute values of the absorption coefficient Ki (;.) And the scattering coefficient Si (λ) of each substance are unknown, if the relative values of the absorption coefficient (λ) and the scattering coefficient Si (λ) of each substance are known, From Equations 1 and 2, the correct spectral reflectance R (Λ) can be determined. Since the preparation of samples for the absolute method is quite difficult and requires complicated work, the relative method is usually used.
従来の相対法によって各物質の吸収係数 Ki (λ ) と散乱係数 Si (λ ) とを求 める際には、 以下のようにして、 基準となる白色顔料の吸収係数 Kw (Λ) と散 乱係数 Sw を決定する。 まず、 被着色物に白色顔料のみを混合した混合物を作成 して、 その分光反射率 R (λ ) を測定する。 従来の相対法では、 被着色物を無色 透明とみなして、 被着色物の吸収係数と散乱係数はともに 0であると仮定してい る。 従って、 被着色物と白色顔料のみを混合した混合物に関しては、 上記の数式 1で与えられる吸収係数 ΚΜ (λ ) と散乱係数 S„ (λ ) は、 白色顔料の散乱係数 Sw と吸収係数 Kw (λ ) にそれぞれ等しい。 また、 上記の数式 2において S„ =SW = 1と仮定すると、 分光反射率 R (/) の測定値から白色顔料の吸収係数 Kw (λ ) を求めることができる。 When determining the absorption coefficient Ki (λ) and the scattering coefficient Si (λ) of each substance by the conventional relative method, the absorption coefficient K w (Λ) of the reference white pigment is calculated as follows. to determine the scattering coefficient S w. First, a mixture is prepared by mixing only a white pigment with an object to be colored, and its spectral reflectance R (λ) is measured. In the conventional relative method, the object to be colored is regarded as colorless and transparent, and it is assumed that both the absorption coefficient and the scattering coefficient of the object to be colored are zero. Thus, for a mixture obtained by mixing only the color bodies and the white pigment, the absorption coefficient given by Equation 1 above kappa Micromax (lambda) and the scattering coefficient S "(lambda) is the scattering coefficient of the white pigment S w and absorption coefficient K w (λ) equal to, respectively. also, obtaining assuming in equation 2 above S "= S W = 1 and the absorption coefficient of the white pigment from the measurements of the spectral reflectance R (/) K w a (lambda) be able to.
上述のように、 従来の相対法では、 被着色物を無色透明とみなして、 その吸収 係数と散乱係数を 0と仮定していた。 ところが、 現実には陶磁器等をはじめ, 被着色物 (陶磁器の場合は釉薬層) が 無色透明でない場合が多い。 白色顔料の調合率が多い場合には、 被着色物が無色 透明であるとみなすことによる誤差は小さいが、 白色顔料の調合率が少ない場合 にはその誤差が無視できない程度に大きくなる。 このような誤差を回避するため には、 上記数式 1において、 無色透明でない被着色物の吸収係数や散乱係数を考 慮にいれる必要がある。 しかし、 被着色物の吸収係数や散乱係数を求めるには、 被着色物単体の薄板を作成するなどの煩雑な作業を要する。 特に、 陶器の釉のよ うに、 単体で薄板を作成することが困難な被着色の場合には、 無色透明でない被 着色物の吸収係数と散乱係数を求めることは困難であった。 As described above, in the conventional relative method, an object to be colored is regarded as colorless and transparent, and its absorption coefficient and scattering coefficient are assumed to be zero. However, in reality, in many cases, the objects to be colored (glaze layers in the case of ceramics), such as ceramics, are not colorless and transparent. When the blending ratio of the white pigment is large, the error caused by assuming that the object to be colored is colorless and transparent is small, but when the blending ratio of the white pigment is small, the error becomes so large that it cannot be ignored. In order to avoid such errors, it is necessary to take into account the absorption coefficient and scattering coefficient of the colorless and non-transparent object in Equation 1 above. However, in order to obtain the absorption coefficient and the scattering coefficient of an object to be colored, a complicated operation such as creating a thin plate of the object to be colored is required. In particular, in the case of coloring, such as ceramic glaze, in which it is difficult to make a thin plate by itself, it was difficult to determine the absorption coefficient and scattering coefficient of a coloring object that is not colorless and transparent.
この発明の第 1の目的は、 無色透明でない被着色物の吸収係数や散乱係数を考 慮してコンピュータカラーマッチングを行なうことにある。  A first object of the present invention is to perform computer color matching in consideration of an absorption coefficient and a scattering coefficient of an object to be colored that is not colorless and transparent.
この発明の第 2の目的は、 コンピュータカラ一マッチングにおける予測誤差を 減少させることにある。  A second object of the present invention is to reduce a prediction error in computer color matching.
この発明の第 3の目的は、 コンピュータカラ一マッチングを利用して, 調合済 み着色剤の有効利用を図りつつ、 その再調合を簡略化することにある。  A third object of the present invention is to make effective use of the prepared colorant by using computer color matching and to simplify re-formulation.
[発明の開示]  [Disclosure of the Invention]
この発明は, コンピュータカラ一マッチングによつて着色剤の調合割合の予測 または混合物の色予測を行なう方法に関する。 このコンピュータカラ一マツチン グ方法は,  The present invention relates to a method for predicting a mixing ratio of a colorant or a color of a mixture by computer color matching. This computer color matching method is
無色透明でない被着色物に白色着色剤を混合した第 1の混合物の散乱係数 S w_ ' を、 前記白色着色剤の調合率 Cw に依存した関数 f ( CK ) として準備するェ 程と、 Preparing a scattering coefficient S w ′ of a first mixture obtained by mixing a white colorant with a colorless and transparent material as a function f (C K ) depending on the blending ratio C w of the white colorant; ,
前記散乱係数 S w' を基準として、 前記第〗の混合物の吸収係数 K w, を前記調 合率 Cw に依存する形式で求める工程と、 Obtaining an absorption coefficient K w , of the second mixture, based on the scattering coefficient S w ′ in a form depending on the mixing ratio C w ;
前記散乱係数 S w' を基準として、 白色でない有色着色剤の吸収係数 K p と散 乱係数 S p とを前記有色着色剤の調合率 Cp に依存する形式で求める工程と、 所望の色を有する混合物を調整するための着色剤の調合割合、 または、 所定の 調合割合で生成される混合物の色を、 前記吸収係数 Kw' , p および前記散乱 係数 Sw' , Sp を用いたコンピュータカラーマッチングを行なうことによって 求める工程と、 The basis of the scattering coefficient S w ', a step of determining a format which depends the turbulent coefficient dispersion and absorption coefficient K p of the colored colorant not white S p in Formulation ratio C p of the colored colorant, a desired color Mixing ratio of the colorant to adjust the mixture having The color of the mixture produced in formulation ratio, the absorption coefficient K w ', p, and the scattering coefficient S w', a step of determining by performing computer color matching using S p,
を備える。 Is provided.
上記方法では、 無色透明でない被着色物に白色着色剤を混合した第】の混合物 の散乱係数 sw, を関数 f (Cw ) で表現し、 この散乱係数 sw, を基準として用 いるので、 散乱係数 sw' に無色透明でない被着色物の影響を舍めることができ る。 In the above method, the scattering coefficient s w , of the second mixture obtained by mixing the white colorant with the non-colorless and transparent coloring object is expressed by a function f (C w ), and this scattering coefficient s w , is used as a reference. In addition, the influence of the non-colorless and transparent object on the scattering coefficient s w 'can be suppressed.
好ましい実施例においては、  In a preferred embodiment,
前記第 1の混合物の散乱係数 sw' を準備する工程は、 Preparing a scattering coefficient s w ′ of the first mixture,
(a) 前記被着色物に前記白色着色剤を混合して、 前記白色着色剤の調合率 Cw が異なる複数個の第 1の混合物を作成するとともに、 前記複数個の第 1の混合物 の分光反射率をそれぞれ測定する工程と、 (a) said mixing the white colorant to be colored object, the spectral of the with Formulation ratio C w of the white colorant to create a plurality of different first mixture, the first mixture of said plurality Measuring the reflectivity,
( b ) 前記被着色物に前記有色着色物を混合した第 2の混合物を作成するととも に、 前記第 2の混合物の分光反射率を測定する工程と、  (b) a step of preparing a second mixture in which the colored object is mixed with the colored object, and measuring a spectral reflectance of the second mixture;
(c) 前記被着色物に前記白色着色剤と前記有色着色剤とを混合して、 前記有色 着色剤の調合率 Cp が異なる複数個の第 3の混合物を作成するとともに、 前記複 数個の第 3の混合物の分光反射率をそれぞれ測定する工程と、 (c) said mixing with the white colorant to be colored object and the colored colorant, along with Formulation ratio C p of the colored colorant to create a plurality of different third mixture, several said double Measuring the spectral reflectances of the third mixture of
(d) 前記被着色物に前記白色着色剤と前記有色着色剤とを混合して、 前記第 3 の混合物とは調合割合が異なる第 4の混合物を作成するとともに、 前記第 4の混 合物の分光反射率を測定する工程と、  (d) mixing the white colorant and the colored colorant with the object to be colored to form a fourth mixture having a different mixing ratio from the third mixture, and the fourth mixture. Measuring the spectral reflectance of the
(e) 前記複数個の第】の混合物の散乱係数 S„' を前記白色着色剤の調合率 Cw の関数 f (Cw ) によって表わすとともに、 前記関数 f (Cw ) に舍まれる係数 の値を仮決定する工程と、 (e) The scattering coefficient S „′ of the plurality of the second mixtures is represented by a function f (C w ) of the mixing ratio C w of the white colorant, and the coefficient f (C w ) Temporarily determining the value of
(f ) 前記複数個の第 1の混合物の分光反射率の測定値と前記関数 f (Cw ) と を用いて、 前記調合率 Cw に依存した形式で前記第 1の混合物の吸収係数 Kw' を求める工程と、 (g) 前記第 2の混合物の分光反射率と、 前記複数個の第 3の混合物の分光反射 率と、 前記関数 f (Cw ) と、 前記吸収係数 Kw' とを用いて、 前記調合率 Cp に依存した形式で前記有色着色剤の吸収係数 と散乱係数 とをそれぞれ 求める工程と、 (f) Using the measured values of the spectral reflectances of the plurality of first mixtures and the function f (C w ), the absorption coefficient K of the first mixture in a form dependent on the blending rate C w w ' (g) using the spectral reflectance of the second mixture, the spectral reflectances of the plurality of third mixtures, the function f (C w ), and the absorption coefficient K w ′, Determining an absorption coefficient and a scattering coefficient of the colored colorant in a format dependent on the rate C p ;
(h) 前記吸収係数 Kw, , KP および前記散乱係数 Sw, , SP を用いて、 前記 第 4の混合物に関するコンピュータカラーマッチングを行なうとともに、 前記コ ンピュータカラーマッチングによる予測精度を向上させるように前記関数 f (C w ) に舍まれる係数を修正する工程と、 (h) using the absorption coefficients K w , K p and the scattering coefficients S w , S p to perform computer color matching on the fourth mixture and improve prediction accuracy by the computer color matching Correcting the coefficient included in the function f (C w ),
( i ) 前記工程 (f) ないし (h) を繰り返すことによって前記関数 ί (Cw ) に舍まれる係数を決定する 1:程と、 (i) determining the coefficients involved in the function ί (C w ) by repeating the steps (f) to (h);
を備える。 Is provided.
こうして関数 f (Cw ) の係数を決定するようにすれば、 コンピュータカラー マッチングによる予測精度が高くなるように関数 f (Cw ) の係数を求めること ができる。 By determining the coefficients of the function f (C w ) in this way, the coefficients of the function f (C w ) can be obtained so that the prediction accuracy by computer color matching is improved.
前記関数 f (cw ) が定数 a i と係数 SB とを含む次の式で与えられるように することが好ましい。 : Wherein the function f (c w) it is preferable to be given by the following equation including the constants ai and coefficient S B. :
f (Cw ) = (CK + a, · SB ) Z (Cw + 3l ) 。 f (C w ) = (C K + a, · S B ) Z (C w + 3l ).
上記の関数 f (CH ) の形式は、 白色着色剤単独の散乱係数を 1と仮定した場 合に理論的に得られるものなので、 散乱係数 Sw' を表わす関数として適切なも のである。 また、 1つの係数 SB を舍むだけの簡単な形式を有しているので、 係 数 SB を決定するのが容易である。 The form of the above function f (C H ) is theoretically obtained assuming that the scattering coefficient of the white colorant alone is 1, and is therefore appropriate as a function representing the scattering coefficient S w '. Also, since one coefficient S B have a simple form only舍Mu, it is easy to determine the engagement number S B.
この発明の他の局面によれば, コンピュータカラーマッチング方法は、 (a) 複数の着色剤を混合して、 調合率が互いに異なる複数のサンプルを準備するェ稈 と、 (b) 前記複数のサンプルの分光反射率をそれぞれ測定するとともに、 前記 分光反射率の測定値から、 前記複数のサンプルのそれぞれの色を表わす所定の表 色系の座標値の実測値を求める工程と、 (c) 前記複数のサンプルのそれぞれに 関して、 前記表色系の座標値の予測誤差を算出する工程と、 (d) 前記複数のサ ンプルに関する前記表色系の座標値と前記予測誤差との関係を、 所定の誤差補正 法で分析する工程と、 (e ) 前記誤差補正法を用いてコンピュータカラーマッチ ングの目標値または予測値を補正しつつ、 新たな混合物の着色剤の調合割合の予 測または混合物の色予測をコンピュータカラ一マッチングにより行なう工程と、 を備える。 According to another aspect of the present invention, a computer color matching method includes: (a) mixing a plurality of colorants to prepare a plurality of samples having different mixing ratios; and (b) the plurality of samples. Measuring the spectral reflectances of the plurality of samples, and obtaining actual measured values of coordinate values of a predetermined color system representing respective colors of the plurality of samples from the measured values of the spectral reflectances; Calculating, for each of the samples, a prediction error of the coordinate values of the color system; and Analyzing the relationship between the coordinate values of the color system with respect to the sample and the prediction error by a predetermined error correction method; and (e) calculating a target value or a prediction value of computer color matching using the error correction method. Predicting the blending ratio of the colorant in the new mixture or predicting the color of the mixture by computer color matching while making corrections.
複数のサンプルに関する所定の表色系の座標値とその予測誤差とを所定の誤差 補正法で分析し、 その誤差補正法によってコンピュータカラーマッチングの目標 値または予測値を補正しつつ予測を行なうので、 各成分の吸収係数 Ki と散乱係 数 S Α とを補正することなく予測誤差を減少させることができる。 Since the coordinate values of the predetermined color system and the prediction errors thereof for a plurality of samples are analyzed by a predetermined error correction method, and the prediction is performed while correcting the target value or the predicted value of the computer color matching by the error correction method. it is possible to reduce the prediction error without correcting the absorption coefficient Ki and scattering coefficient number S Alpha of each component.
好ましい実施例においては、 前記工程 (d ) は、 前記複数のサンプルに関する 前記表色系の座標値と前記予測誤差との関係を二ユーラルネットワークに学習さ せる工程、 を含み、 前記工程 (e ) は、 学習済みのニューラルネットワークを用 いてコンピュータカラ一マッチングによる予測を行なう工程を舍む。  In a preferred embodiment, the step (d) includes a step of causing a dual neural network to learn a relationship between the coordinate values of the color system and the prediction error for the plurality of samples, and the step (e). ) Shows a process of performing prediction by computer color matching using a trained neural network.
ニューラルネットワークを用いて予測誤差を補正するようにすれば、 数多くの サンプルを学習させることによって、 予測誤差を減少させることができる。 また、 前記ニューラルネットワークは、 3つのニューロンで構成される入力層 と、 複数のニューロンを舍む中間層と、 3つのニューロンで構成される出力層と、 で構成される三層の階餍構造を有することが好ましい。  If the prediction error is corrected using a neural network, the prediction error can be reduced by learning a large number of samples. The neural network has a three-layer hierarchical structure including an input layer including three neurons, an intermediate layer including a plurality of neurons, and an output layer including three neurons. It is preferred to have.
この発明のさらに他の局面によれば, コンピュータカラーマッチング方法は、 所望の色を有する目標混合物の着色剤の調合割合を予測する方法であって、 According to yet another aspect of the present invention, a computer color matching method is a method for predicting a mixing ratio of a colorant in a target mixture having a desired color,
( a ) 着色剤の調合率が既知で前記所望の色に近い色を有する近接色サンプルに ついて所定の表色系の座標の実測値を求める工程と、 (a) obtaining an actual measurement value of coordinates of a predetermined color system for an adjacent color sample having a colorant formulation known and having a color close to the desired color;
( b ) 前記近接色サンプルの既知の調合割合から、 前記近接色サンプルの色を表 わす前記表色系の座標の計算値を求め、 前記実測値と前記計算使から計算誤差を 求める工程と、  (b) obtaining a calculated value of the coordinates of the color system representing the color of the adjacent color sample from the known blending ratio of the adjacent color sample, and obtaining a calculation error from the actual measurement value and the calculator.
( c ) ¾UH目標混合物の色に対する ii己表色系の座標の目標値を設定する工程と、 (c) setting a target value of ii self-color system coordinates for the color of the ¾UH target mixture;
( d ) 前記計算誤差を用いて前記目標値を補正し、 補正後の目標値を用いてコン ピュータカラーマツチングを実行することによって、 前記目標混合物の着色剤の 調合割合を予測する工程と、 (d) correcting the target value by using the calculation error, and executing computer color matching using the corrected target value, whereby the colorant of the target mixture is Estimating the blending ratio;
を備える。 Is provided.
近接色サンプルの表色系の座標値の計算誤差によって目標混合物の表色系の座 標の目標彼を補正し、 補正後の目標値を用いてコンピュータカラーマッチングを 行なうので、 表色系の座標値の計算誤差を低減した条件下でコンピュータカラー マッチングを実行できる。  The target colorimetric coordinate of the target mixture is corrected by the calculation error of the coordinate value of the colorimetric system of the adjacent color sample, and computer color matching is performed using the corrected target value. Computer color matching can be performed under conditions that reduce value calculation errors.
好ましい実施例によれば、 前記工程 (a ) は、 複数のサンプルに関して、 着色 剤の調合割合と、 前記表色系の座標の実測値とを含むデータベースから、 前記目 標混合物との色差が最小となるサンプルを選択することによつて前記近接色サン プルを検索する工程を舍む。  According to a preferred embodiment, in the step (a), for a plurality of samples, a color difference from the target mixture is minimized from a database including a mixing ratio of a coloring agent and an actual measurement value of coordinates of the color system. The step of searching for the adjacent color sample by selecting the sample to be performed is provided.
データベースから色差が最小のサンプルを近接色サンプルとして検索するよう にすれば、 任意の色を有する目標混合物に関して、 コンピュータカラーマツチン グの予測誤差を低減できる。  By retrieving the sample having the smallest color difference from the database as the closest color sample, it is possible to reduce the prediction error of computer color matching for a target mixture having an arbitrary color.
この発明のさらに他の局面によれば, コンピュータカラ一マッチング方法は、 複数の着色剤を調合した調合物が所望の目標色に近似した色を呈するように該 複数の着色剤についての調合割合を求める方法であって、  According to yet another aspect of the present invention, a computer color matching method includes: adjusting a compounding ratio of a plurality of colorants so that the mixture obtained by mixing the plurality of colorants exhibits a color similar to a desired target color. Is the method of seeking
( a ) 前記目標色を呈する調合物見本について、 所定の表色系での色評価値の 実測値を求める工程と、  (a) obtaining a measured value of a color evaluation value in a predetermined color system for a preparation sample exhibiting the target color;
( b ) 既知の調合割合で前記着色剤が調合された〗次調合物について、 前記所 定の表色系での色評価値の実測値を求める工程と、  (b) a step of obtaining an actual measurement value of a color evaluation value in the predetermined color system, for a secondary preparation in which the colorant is prepared at a known preparation ratio;
( c ) 前記 1次調合物についての前記既知の調合割合に基づいて、 前記 1次調 合物の呈する色の前記所定の表色系での色評価値の計算値を求める工程と、 (c) obtaining a calculated value of a color evaluation value of the color represented by the primary formulation in the predetermined color system, based on the known blending ratio of the primary formulation;
( d ) 前記】次調合物に前記着色剤を増量補正したと仮定した着色剤増量調合 物についての前記所定の表色系での色評価値の計算値を求め、 前記 1次調合物か ら前記着色剤增量調合物への前記色評価値の計算値の変化量を求める工程と、(d) Calculating the color evaluation value in the predetermined color system for the colorant-enhanced formulation assuming that the colorant in the next formulation has been corrected for the increased amount of the colorant, and obtaining the calculated value from the primary formulation Determining the amount of change in the calculated value of the color evaluation value to the colorant amount formulation,
( e ) 前記調合物見本と前記 1次調合物との前記色評価値の実測値の差が所定 範囲で一致するように、 前記色評価値の計算値の変化量に基づいて前記着色剤の それぞれの増量補正量を算出する丁-程と、 を備える。 (e) the colorant based on the amount of change in the calculated value of the color evaluation value so that the difference between the measured value of the color evaluation value of the sample of the formulation and the primary formulation is equal within a predetermined range. And a step for calculating each amount of increase correction.
ヒ記のコンピュータカラーマッチング方法では、 既知の調合割合の 1次調合物 に着色剤を仮に増量補正した場合、 その際の色評価値の計算値の変化量を求める。 そして、 この色評価値の計算ィ ¾の変化量に基づいてそれぞれの着色剤についての 増量補正量を算出し、 調合物見本と 1次調合物との色評価値の実測値の差を所定 範囲で一致させる。 このため、 求めた增暈補正量だけそれぞれの着色剤を 1次調 合物に実際に追加調合すれば、 調合物見本の目標色若しくはこれに近似した色を 呈する調合物に、 1次調合物を再調合できる。 従って、 本発明のコンピュータ力 ラーマッチング方法によれば、 着色剤の除去を意味する負の補正量を求めること がないので、 着色剤が調合済みの調合物 ( 1次調合物) の廃棄が不要となり、 既 存の調合物を有効に利用することができる。 また、 この際に 1次調合物を既存の 調合物ではなく新たに試験的に調合するにしても、 その調合に技術者が一回限り 関与すればよく、 しかもその際に、 技術者の長年の勘や経験を要しないので、 調 合物の再調合を簡略化することができる。  According to the computer color matching method described in Hi, when the colorant is temporarily increased and corrected in a primary formulation having a known blending ratio, the amount of change in the calculated color evaluation value at that time is determined. Then, based on the amount of change of the color evaluation value calculation amount, the amount of increase correction for each colorant is calculated, and the difference between the actually measured value of the color evaluation value of the preparation sample and that of the primary preparation is determined within a predetermined range. To match. For this reason, if each colorant is actually added to the primary formulation by the calculated halo correction amount, the primary formulation will be added to the target color of the sample of the formulation or a formulation exhibiting a color similar to the target color. Can be reconstituted. Therefore, according to the computer color matching method of the present invention, there is no need to obtain a negative correction amount meaning removal of the colorant, so that it is not necessary to dispose of the preparation in which the colorant has been prepared (primary preparation). The existing formulation can be used effectively. In this case, even if the primary formulation is newly prepared on a trial basis instead of the existing formulation, the technician need only be involved once in the formulation, and at that time, Since no intuition or experience is required, re-mixing of the composition can be simplified.
好ましい実施例においては、  In a preferred embodiment,
前記工程 (d ) は、 前記〗次調合物における前記着色剤の調合量に比べて微量 の量の前記着色剤を前記 1次調合物に増量補正したと仮定した場合について、 前 記色評価値の計算値の変化量を求める工程を含む。  In the step (d), the color evaluation value is calculated based on the assumption that a small amount of the coloring agent is corrected to be added to the primary formulation compared to the blending amount of the coloring agent in the secondary formulation. Calculating the amount of change in the calculated value.
このコンピュータカラーマツチング方法では、 色評価値の計算値の変化量を求 めるために行なう着色剤の仮の増量補正を微量単位で行なって、 微小単位の変化 量を求めることができる。 よって、 それぞれの着色剤の増量補正量を高い精度で 求めることを通して、 目標色によく一致した色を呈することができる調合物を得 ることができる。  In this computer color matching method, a temporary increase correction of the colorant, which is performed in order to obtain a change amount of the calculated value of the color evaluation value, is performed in a minute unit, and a change amount in a minute unit can be obtained. Thus, by obtaining the amount of correction for the amount of each colorant with high accuracy, it is possible to obtain a composition that can exhibit a color that is in good agreement with the target color.
また、 前記工程 (e ) は、 前記着色剤の増量に伴う派生費用を表わす費用関数 を用いた線形計 ί面法にて、 前記着色剤のそれぞれについての最小の増量補正量を 算出する工程を舍むことが好ましい。  Further, the step (e) includes a step of calculating the minimum amount of increase correction for each of the colorants by a linear measurement surface method using a cost function representing a derivation cost associated with the increase in the amount of the colorant. It is preferable to settle.
このコンピュータカラーマツチング方法では、 それぞれの着色剤を実際に追加 増量する際の最小の増量補正量を算出できるので、 着色剤の使用量の低減を通し て派生費用の最小化を図ることができ、 コストを低減することができる。 In this computer color matching method, since the minimum amount of increase correction when each colorant is actually added is calculated, the amount of colorant used can be reduced. As a result, derivation costs can be minimized and costs can be reduced.
この発明は, また, コンピュータカラ一マッチングによって着色剤の調合割合 の予測または混合物の色予測を行なう装置に向けられている。 この装置は、 無色透明でない被着色物に白色着色剤を混合した第〗の混合物の散乱係数 S w_The present invention is also directed to an apparatus for predicting a mixing ratio of a colorant or a color of a mixture by computer color matching. This device has a scattering coefficient S w _ of a second mixture in which a white colorant is mixed with a colorless and transparent object to be colored.
' を、 前記白色着色剤の調合率 Cw に依存した関数 f ( Cw ) として作成する手 段と、 'As a function f (C w ) depending on the blending ratio C w of the white colorant,
前記散乱係数 s w, を基準として、 前記第〗の混合物の吸収係数 κ„, を前記調 合率 cw に依存する形式で求める手段と、 Means for determining an absorption coefficient κ „, of the second mixture in a form depending on the mixing rate c w , based on the scattering coefficient s w ,
前記散乱係数 s w' を基準として、 白色でない有色着色剤の吸収係数 K P と散 乱係数 s P とを前記有色着色剤の調合率 cp に依存する形式で求める孚段と、 所望の色を有する混合物を調整するための若色剤の調合割合、 または、 所定の 調合割合で生成される混合物の色を、 前記吸収係数 Kw' , K P および前記散乱 係数 s w' , S P を用いたコンピュータカラーマッチングを行なうことによって 求める手段と、 With reference to the scattering coefficient s w ′, the absorption coefficient K P and the scattering coefficient s P of the non-white colored colorant are determined in a form depending on the mixing ratio c p of the colored colorant, and a desired color is obtained. The blending ratio of a young agent for preparing a mixture having the following formulas, or the color of the mixture produced at a predetermined blending ratio, is calculated using the absorption coefficients K w ′, K P and the scattering coefficients s w ′, S P Means to obtain by performing the computer color matching used;
を備える。 Is provided.
この発明の他の局面によれば, コンピュータカラ一マツチング装置は、 複数の着色剤を混合することによつて作成された調合率が互いに異なる複数の サンプルの分光反射率をそれぞれ測定する手段と,  According to another aspect of the present invention, a computer color matching device comprises: means for measuring spectral reflectances of a plurality of samples having different mixing ratios formed by mixing a plurality of colorants;
前記分光反射率の測定値から、 前記複数のサンプルのそれぞれの色を表わす所 定の表色系の座標値の実測値を求める手段と、  Means for obtaining, from the measured values of the spectral reflectance, actual measured values of coordinate values of a predetermined color system representing respective colors of the plurality of samples;
前記複数のサンプルのそれぞれに関して、 前記表色系の座標値の予測誤差を算 出する手段と、  Means for calculating a prediction error of the coordinate value of the color system for each of the plurality of samples;
前記複数のサンプルに関する前記表色系の座標値と前記予測誤差との関係を、 所定の誤差補正法で分析する手段と、  Means for analyzing the relationship between the coordinate values of the color system for the plurality of samples and the prediction error, using a predetermined error correction method,
前記誤差補正法を用いてコンピュータカラーマッチングの目標値または予測値 を補正しつつ、 新たな混合物の着色剤の調合割合の予測または混合物の色予測を コンピュータカラーマッチングにより行なう手段と、 を備える。 Means for correcting the target value or the predicted value of the computer color matching using the error correction method, and performing the prediction of the mixing ratio of the colorant in the new mixture or the color prediction of the mixture by computer color matching, Is provided.
この発明のさらに他の局面によれば, コンピュータカラ一マッチング装置は、 着色剤の調合率が既知で前記所望の色に近い色を有する近接色サンプルについ て所定の表色系の座標の実測値を求める手段と、  According to still another aspect of the present invention, a computer color matching apparatus includes: an actual measurement value of coordinates of a predetermined color system for a proximity color sample having a colorant mixing ratio known and having a color close to the desired color; Means to determine
前記近接色サンプルの既知の調合割合から、 前記近接色サンプルの色を表わす 前記表色系の座標の計算値を求め、 ii記実測値と前記計算値から計算誤差を求め る丰段と、  Calculating a calculated value of the coordinates of the color system representing the color of the adjacent color sample from the known blending ratio of the adjacent color sample, and ii obtaining a calculation error from the actually measured value and the calculated value;
前記目標混合物の色に対する前記表色系の座標の目標値を設定する手段と、 前記計算誤差を用いて前記目標値を補正し、 補正後の目標値を用いてコン ピュータカラーマッチングを実行することによって、 前記目標混合物の着色剤の 調合割合を予測する手段と、  Means for setting a target value of the coordinates of the color system for the color of the target mixture; correcting the target value using the calculation error; and executing computer color matching using the corrected target value. Means for predicting the blending ratio of the colorant in the target mixture,
を備える。 Is provided.
この発明の他の局面によれば, コンピュータカラーマッチング装置は, 複数の 着色剤を調合した調合物が所望の目標色に近似した色を呈するように該複数の着 色剤についての調合割合を求めるものである。 このコンピュータカラーマツチン グ装置は、  According to another aspect of the present invention, a computer color matching device calculates a blending ratio of a plurality of coloring agents so that the formulation obtained by blending the plurality of coloring agents exhibits a color similar to a desired target color. Things. This computer color matching device
前記目標色を呈する調合物見本について、 所定の表色系での色評価値の実測値 を求める手段と、  Means for obtaining an actual measurement value of a color evaluation value in a predetermined color system for a preparation sample exhibiting the target color;
既知の調合割合で前記着色剤が調合された】次調合物について、 前記所定の表 色系での色評価値の実測俯を求める手段と、  Means for obtaining the actually measured depression of the color evaluation value in the predetermined color system for the next preparation, wherein the colorant is prepared at a known preparation ratio;
前記〗次調合物についての前記既知の調合割合に基づいて、 前記 1次調合物の 呈する色の前記所定の表色系での色評価値の計算値を求める丰段と、  A step of obtaining a calculated value of a color evaluation value of the color represented by the primary formulation in the predetermined color system based on the known blending ratio of the primary formulation;
前記 1次調合物に前記着色剤を増量補正したと仮定した着色剤増量調合物につ いての前記所定の表色系での色評価値の計算値を求め、 前記 1次調合物から前記 着色剤増量調合物への前記色評価値の計算値の変化量を求める手段と、  Calculating the color evaluation value in the predetermined color system for the colorant-enhanced formulation assuming that the colorant has been increased in the primary formulation and calculating the coloration from the primary formulation Means for determining the amount of change in the calculated value of the color evaluation value to the agent extension formulation,
前記調合物見本と前記 1次調合物との前記色評価値の実測値の差が所定範囲で 一致するように、 前記色評価値の計算値の変化量に基づいて前記着色剤のそれぞ れの増量補正量を算出する手段と、 を傰える。 Each of the colorants based on the amount of change in the calculated value of the color evaluation value so that the difference between the measured value of the color evaluation value of the preparation sample and that of the primary preparation matches within a predetermined range. Means for calculating the increase correction amount.
この発明は, さらに, コンピュータカラーマッチングに使用される着色剤の吸 収係数と散乱係数を決定する方法に向けられている。 この方法は、  The invention is further directed to a method for determining the absorption and scattering coefficients of a colorant used in computer color matching. This method
無色透明でない被着色物に白色着色剤を混合した第 1の混合物の散乱係数 sw_ ' と吸収係数 κ„' を決定する工程と、 Determining a scattering coefficient s w _ 'and an absorption coefficient κ „' of a first mixture obtained by mixing a white colorant with a colorless and transparent object;
前記散乱係数 sw' と吸収係数 Kw, の少なくとも一方を基準として、 白色でな い有色着色剤の吸収係数 κρ と散乱係数 Sp の相対値をそれぞれ決定する工程 と、 を備える。 Determining the relative values of the absorption coefficient κ ρ and the scattering coefficient Sp of the non-white colored colorant on the basis of at least one of the scattering coefficient s w ′ and the absorption coefficient K w , respectively.
好ましい実施例によれば, 前記第〗の混合物の散乱係数 sw' 力 前記白色着 色剤の調合率 cwと定数 3 l と係数 SB とを舍む次の関係で与えられる: According to a preferred embodiment, the scattering coefficient s w ′ of the second mixture is given by the following relationship involving the mixing ratio c w of the white colorant, a constant 3 l and a coefficient S B :
Sw' =(CW +a, -SB )/(C„ +a】 ). S w '= (C W + a, -S B ) / (C „+ a ]).
他の好ましい実施例によれば, 前記第〗の混合物の散乱係数 Sw' 力, 前記白 色着色剤の調合率 Cwと定数 a, b, d, e ,CWUを舍む次の関係で与えられる:According to another preferred embodiment, the scattering coefficient S w 'of the second mixture, the mixing ratio C w of the white colorant and the following relations a, a, b, d, e, C WU Given by:
Sw, - a-Cw + b (Cw。 ≤ Cw); S w ,-aC w + b (C w . ≤ C w );
Sw, = d-Cw + e (Cw < Cw0). S w , = dC w + e (C w <C w0 ).
この発明は, また, コンピュータカラーマッチングを利用して作成された混合 釉に向けられている。 この混合釉は,  The present invention is also directed to mixed glazes made using computer color matching. This mixed glaze
無色透明でないベース釉に白色着色剤を混合した第】の混合釉の散乱係数 Sw_ ' と吸収係数 Kw' を決定し、 Determine the scattering coefficient S w _ 'and the absorption coefficient K w ' of the mixed glaze in which a white colorant is mixed with a colorless base glaze,
前記散乱係数 Sw, と吸収係数 Kw' の少なくとも一方を基準として、 白色でな い有色着色剤の吸収係数 Kp と散乱係数 Sp の相対値をそれぞれ決定し、 前記吸収係数 Kw, , KP および前記散乱係数 sw, , SP を用いたコンビユー タカラーマッチングを行なうことによって, 所望の色を有する混合釉を調整する ための着色剤の調合割合、 または、 所定の調合割合で生成される混合釉の色を予 測し、 Based on at least one of the scattering coefficient S w , and the absorption coefficient K w ′, the relative values of the absorption coefficient K p and the scattering coefficient S p of the non-white colored colorant are determined, and the absorption coefficient K w , , K P and the scattering coefficient s w,, by performing Konbiyu Takara over matching using S P, formulation ratio of the colorant to adjust the mixed glaze having a desired color, or, at a predetermined compounding ratio Predict the color of the resulting mixed glaze,
前記予測された調合割合または前記予測された色を有するように前記ベース釉 と前記白色着色剤と前記有色着色剤とを混合する, ことによって生成される。 Mixing the base glaze, the white colorant, and the colored colorant to have the predicted mix proportion or the predicted color; Generated by
この発明は, さらに, コンピュータカラーマッチングを利用して作成された混 合釉を用いて製造された陶磁器にも向けられている。 この陶磁器は、  The present invention is further directed to ceramics manufactured using a mixed glaze made using computer color matching. This ceramic is
無色透明でないベ一ス釉に白色着色剤を混合した第 1の混合釉の散乱係数 S w— ' と吸収係数 Kw' を決定し、 Determine the scattering coefficient S w — 'and the absorption coefficient K w ' of the first mixed glaze in which a white colorant is mixed with a colorless and transparent base glaze,
前記散乱係数 s w' と吸収係数 K w' の少なくとも一方を基準として、 白色でな い有色着色剤の吸収係数 κ ρ と散乱係数 S p の相対値をそれぞれ決定し、 前記吸収係数 Kw, , K p および前記散乱係数 s w, , S p を用いたコンビユー タカラーマツチングを行なうことによって, 所望の色を有する混合釉を調整する ための着色剤の調合割合、 または、 所定の調合割合で生成される混合釉の色を予 測し、 Based on at least one of the scattering coefficient s w ′ and the absorption coefficient K w ′, the relative values of the absorption coefficient κ ρ and the scattering coefficient S p of the non-white colored colorant are determined, and the absorption coefficient K w , , K p and the scattering coefficient s w, by performing Konbiyu Takara over pine quenching with S p, formulation ratio of the colorant to adjust the mixed glaze having a desired color or a predetermined compounding ratio Predict the color of the mixed glaze generated by
前記予測された調合割合または前記予測された色を有するように前記ベース釉 と前記白色着色剤と前記有色着色剤とを混合する, ことによって生成された混合 釉で陶磁器の素地の少なくとも一部が覆われる。  Mixing the base glaze, the white colorant, and the colored colorant to have the predicted blending ratio or the predicted color, wherein at least a portion of the ceramic body is Covered.
この発明の他の局面においては, コンピュータカラーマッチングに使用される 着色剤の吸収係数と散乱係数を決定する方法は、  In another aspect of the present invention, a method for determining an absorption coefficient and a scattering coefficient of a colorant used in computer color matching comprises:
無色透明でない被着色物に基準となる顔料を混合した第〗の混合物の散乱係数 s と吸収係数 Kw' を決定する工程と、 A step of determining a scattering coefficient s and an absorption coefficient K w ′ of a second mixture obtained by mixing a reference pigment with a colorless and non-colored object;
前記散乱係数 S w, と吸収係数 K w, の少なくとも一方を基準として、 有色着色 剤の吸収係数 κ ρ と散乱係数 S P の相対値をそれぞれ決定する工程と、 を備える。 The scattering coefficient S w, the absorption coefficient K w, based on the at least one, and a step of determining the relative value of the absorption coefficient kappa [rho and scattering coefficient S P output colored colorant, respectively, the.
この発明のさらに他の局面によれば, コンピュータカラーマッチングに使用さ れる着色剤の吸収係数と散乱係数を決定する装置は、  According to yet another aspect of the present invention, an apparatus for determining an absorption coefficient and a scattering coefficient of a colorant used in computer color matching comprises:
無色透明でない被着色物に基準となる顔料を混合した第】の混合物の散乱係数 s と吸収係数 Kw' を決定する手段と、 Means for determining the scattering coefficient s and the absorption coefficient K w ′ of the second mixture of the non-colorless and transparent coloring matter mixed with the reference pigment;
前記散乱係数 S w' と吸収係数 Kw' の少なくとも一方を基準として、 有色着色 剤の吸収係数 K P と散乱係数 S p の相対値をそれぞれ決定する:に程と、 を備える。 On the basis of at least one of the scattering coefficient S w ′ and the absorption coefficient K w ′, determine the relative values of the absorption coefficient K P and the scattering coefficient S p of the colorant, respectively: Is provided.
また, この発明は. コンピュータカラーマッチングを利用して作成された混合 釉を用いて製造された便器にも向けられている。  The present invention is also directed to a toilet bowl manufactured using mixed glaze made using computer color matching.
[図面の鲔. な説明] [A description of the drawing]
図 1は, 第 1実施例における処理の全体手順を示すフローチヤ一ト。  FIG. 1 is a flowchart showing the overall processing procedure in the first embodiment.
図 2は, 数式 5で与えられる散乱係数 S w' を示すグラフ。 Fig. 2 is a graph showing the scattering coefficient S w 'given by Eq.
図 3は, ステップ S 1の詳細手順を示すフローチャート。  Fig. 3 is a flowchart showing the detailed procedure of step S1.
図 4は, (ベース釉 +白色顔料) の物性彼を求めるためのサンプルの調合率を 示す説明図。  Fig. 4 is an explanatory diagram showing the mixing ratio of a sample to determine the physical properties of (base glaze + white pigment).
図 5は, サンプルの分光反射率 R (;. ) を示す概念図。  Figure 5 is a conceptual diagram showing the spectral reflectance R (;.) Of the sample.
図 6は, ステップ S 1 4の詳細手順を示すフローチャート。  FIG. 6 is a flowchart showing the detailed procedure of step S14.
図 7は, 顔料物性値決定用サンプルの調合率を示す説明図。  Fig. 7 is an explanatory diagram showing the mixing ratio of the pigment physical property value determination sample.
図 8は, 第 1実施例で用いた検証用サンプルの調合率を示す説明図。  FIG. 8 is an explanatory diagram showing the mixing ratio of the verification sample used in the first embodiment.
図 9は, 吸収係数 Κ Ρ ( Λ ) の調合率 Cpに対する依存性の一例を示すグラフ。 図 1 0は, 第 1実施例における調合率の予測結果を示す説明図。 Figure 9 is a graph showing an example of the dependence of the absorption coefficient Κ Ρ (Λ) on the blending rate C p . FIG. 10 is an explanatory diagram showing a prediction result of a mixing ratio in the first embodiment.
図 1 1は, 第 2実施例における処理の全体手順を示すフローチャート。  FIG. 11 is a flowchart illustrating an overall procedure of a process according to the second embodiment.
図 1 2は, ニューラルネットワークの構成を示す説明図。  Figure 12 is an explanatory diagram showing the configuration of a neural network.
図 1 3は, ステップ S 3 1の詳細手順を示すフローチャート。  FIG. 13 is a flowchart showing the detailed procedure of step S31.
図 1 4は, コンピュータカラーマッチングで予測対象範囲 P Aと複数のサンプ ル M 1〜M 7の三刺激値の分布を示す概念図。  Fig. 14 is a conceptual diagram showing the distribution of tristimulus values in the prediction target range PA and multiple samples M1 to M7 in computer color matching.
図 1 5は, 第 1実施例における 7つのサンプル M 1〜M 7の予測誤差を示す概 念図。  FIG. 15 is a conceptual diagram showing prediction errors of seven samples M1 to M7 in the first embodiment.
図 1 6は, 各サンプルの三刺激値の目標値 (ステップ S 4 1で決定された値) と予測誤差 AMi (Δ Χ, Δ Υ, Δ Ζ) を示す説明図。  Figure 16 is an explanatory diagram showing the target tristimulus values (the values determined in step S41) and the prediction error AMi (ΔΧ, ΔΥ, ΔΖ) for each sample.
図 1 7は, 第 2実施例におけるニューラルネットワークの学習の実証結果を示 す説明図。 図 1 8は, 第 2実施例におけるコンピュータカラ一マッチングの予測の実証結 果を示す説明図。 Figure 17 is an explanatory diagram showing the results of the neural network learning in the second embodiment. Fig. 18 is an explanatory diagram showing the verification results of the prediction of computer color matching in the second embodiment.
図 1 9は, 設計された色 (標準色) と実際に製造される陶器の色のばらつきを 示す X— y色度図。  Fig. 19 is an X-Y chromaticity diagram showing the variation between the designed colors (standard colors) and the colors of the ceramics actually manufactured.
図 2 0は, 第 3実施例における処理の全体手順を示すフ π—チャート。  FIG. 20 is a π-chart showing the overall procedure of the process in the third embodiment.
図 2 1は, ステップ S 5 7におけるコンピュータカラ一マッチングの詳細手順 を示すフローチヤ一ト,  Figure 21 is a flowchart showing the detailed procedure of computer color matching in step S57,
図 2 2は, 第 3実施例において用いた檫準色見本の三刺激値と溢淡限度見本の 三刺激値を示すテーブル。  FIG. 22 is a table showing the tristimulus values of the reference color sample and the tristimulus values of the overflow limit sample used in the third embodiment.
図 2 3は, 第 3実施例による渔淡限度見本の三刺激値の予測結果と比较例と実 調合率とを示すテーブル。  Fig. 23 is a table showing the results of the prediction of tristimulus values of the light limit sample, comparative examples, and the actual mixing ratio according to the third embodiment.
図 2 4は, 各実施例のコンピュータカラ一マッチング方法を実施するための装 置を示すブ口ック図。  Figure 24 is a block diagram showing the equipment for implementing the computer color matching method of each embodiment.
図 2 5は, 第 4実施例のコンビュ一タカラーマッチング方法における処理の全 体手順を示すフローチヤ一ト,  FIG. 25 is a flowchart showing the overall procedure of the process in the computer color matching method of the fourth embodiment.
図 2 6は, 図 2 5のステップ S 7 6の誶細処理を示すフローチヤ一ト。  FIG. 26 is a flowchart showing the detailed processing of step S76 in FIG.
図 2 7は、 ステップ S 7 2で取得した目標色見 につレヽての U激値 (色値) とステップ S 7 3で取得した第 1回目試作釉についての三剌瀲値の対比、 並びに 目標色見本釉, 第 1回目試作釉における各顔料の調合率 (調合率) を表わすテ一 ブル。  Figure 27 shows the comparison between the extreme U value (color value) for the target color cast obtained in step S72 and the value of the tristimulus for the first prototype glaze obtained in step S73, and Table showing the mixing ratio (mixing ratio) of each pigment in the target color sample glaze and the first prototype glaze.
図 2 8は、 図 2 7に掲げる第 1回目試作釉に各顔料をそれぞれ微量だけずつ追 力 U調合した場合の三刺激値の変化率 (微係数) を示すテーブル。  Fig. 28 is a table showing the rate of change (derivative coefficient) of tristimulus values when each pigment is added to the first trial glaze shown in Fig. 27 by a small amount of each pigment.
図 2 9は、 第 4実施例によるコンピュータカラ一マッチング方法による結果を 示すテーブル。  FIG. 29 is a table showing the result of the computer color matching method according to the fourth embodiment.
図 3 0は、 第 4実施例による他の結果を示すテーブル。  FIG. 30 is a table showing another result according to the fourth embodiment.
[発明を実施するための最良の形態]  [Best Mode for Carrying Out the Invention]
A. 第 1実施例: 図 1は、 第】実施例における処理の全体手順を示すフローチャートである。 な お、 この第 1実施例で対象とする混合物は、 陶磁器の素地の表面を覆うための釉 (ゆう) である。 すなわち、 顔料を入れないベ一ス釉 (基礎釉) が被着色物であ り、 このべ一ス釉に顔料を添加した釉がコンピュータカラーマツチングの対象と なる混合物である。 A. First embodiment: FIG. 1 is a flowchart showing the overall procedure of the process in the embodiment. The mixture of interest in the first embodiment is a glaze for covering the surface of the ceramic body. In other words, the base glaze (base glaze) containing no pigment is the object to be colored, and the glaze obtained by adding a pigment to this base glaze is a mixture to be subjected to computer color matching.
なお, 陶磁器において, 白色顔料としては乳濁剤が有効である。 乳濁剤として は, ジルコン等のジルコニウム化合物や, リン酸カルシウム等のリン化合物が用 いられる。  In ceramics, an emulsifier is effective as a white pigment. Emulsifiers include zirconium compounds such as zircon and phosphorus compounds such as calcium phosphate.
ステップ S 〗では、 被着色物 (ベース釉) と白色顔料の混合物の吸収係数 K w_ ' と散乱係数 S w' とを求める。 ここで、 従来の相対法と異なるのは、 被着色物 を無色透明と仮定していない点、 および、 被着色物や白色顔料単独の吸収係数や 散乱係数を求めずに、 その混合物の吸収係数 Kw' と散乱係数 s w' を求める点に ある。 なお、 散乱係数 S w' は、 白色顔料の調合率 Cw に依存した関数 f ( Cw In step S 吸収, the absorption coefficient K w _ 'and the scattering coefficient S w ' of the mixture of the object to be colored (base glaze) and the white pigment are determined. Here, the difference from the conventional relative method is that the material to be colored is not assumed to be colorless and transparent, and the absorption coefficient of the mixture is determined without calculating the absorption coefficient or scattering coefficient of the material to be colored or the white pigment alone. The point is to find K w 'and the scattering coefficient s w '. Incidentally, the scattering coefficient S w 'is a function depending on the compounding ratio C w of the white pigment f (C w
) として求める。 また、 吸収係数 κ„' も、 白色顔料の調合率 cw に依存した形 式で求める。 ). Also, the absorption coefficient κ „'is determined in a form that depends on the white pigment blending ratio c w .
なお、 ステップ S 1において, 白色顔料以外の顔料 (以下、 「有色顔料」 と呼 ぶ) の吸収係数 K p と散乱係数 S p も決定される (ステップ S 2 ) 。 なお、 有 色顔料の吸収係数 と散乱係数 は、 その調合率 cp に依存した形式で求 められる。 Note that, in step S 1, pigment other than white pigment (hereinafter, "colored pigments" and hump) absorption coefficient K p and scattering coefficient S p of is also determined (step S 2). The absorption and scattering coefficients of the chromatic color pigment is determined Me in a format depending on the compounding ratio c p.
ステップ S 3では、 ステップ S l, S 2で求めた吸収係数と散乱係数に基づい て、 上述した数式 1および数式 2を用いてコンピュータカラーマッチングを実行 し、 混合物の色予測や調合割合の予測を行なう。  In step S3, computer color matching is performed using the above-described equations 1 and 2 based on the absorption coefficient and the scattering coefficient obtained in steps S1 and S2, and the color prediction of the mixture and the prediction of the blending ratio are performed. Do.
ここで、 ステップ S 1の詳細手順を説明する前に、 ベース釉と白色顔料の混合 物の散乱係数 S w' を表現するための関数 f ( Cw ) について考察する。 ベース 釉と白色顔料の混合物については、 数式 1を次の数式 3のように書き換えること ができる。 Here, before describing the detailed procedure of step S1, a function f (C w ) for expressing the scattering coefficient S w ′ of the mixture of the base glaze and the white pigment will be considered. For a mixture of base glaze and white pigment, Equation 1 can be rewritten as Equation 3 below.
(C W + CB)KW ' = CwKw + RKB . . . ( 3a ) (C W + C B ) K W '= C w Kw + RKB... (3a)
(C、V + CB)SW ' = CwSw +し BSR — ( 3b ) ここで、 Kw, , Sw' は混合物の吸収係数と散乱係数、 Kw , Sw は白色顔料 単独の吸収係数と散乱係数、 KB , SB はベース釉単独の吸収係数と散乱係数、 cw は白色顔料の調合率、 cB はべ一ス釉の調合率である。 なお、 調合率 cw, cB は正確には体積率で表現されるべきであるが、 重量率で表現してもその誤差 は通常無視できる程度である。 (C, V + C B ) S W '= C w Sw + BSR — (3b) Here, K w,, S w 'is the absorption coefficient and scattering coefficient of the mixture, K w, S w is the absorption coefficient and scattering coefficient of the white pigment alone, K B, S B is the scattering and absorption coefficient of the base glaze alone coefficient , C w is the mixing ratio of the white pigment, and c B is the mixing ratio of the base glaze. It should be noted that the mixing ratios c w and c B should be expressed accurately by volume ratio, but even if expressed by weight ratio, the error is usually negligible.
数式 3を変形すると、 混合物の散乱係数 Sw' は次の数式 4で与えられる。 , _ Cw SW+レ RSR . ΛBy transforming Equation 3, the scattering coefficient S w 'of the mixture is given by Equation 4 below. , _ Cw S W + Les RSR. Λ,
Sw- cw+cB ·'·(4 ) この発明では相対法を用いるので、 白色顔料の調合率 cwによらず散乱係数 sw Sw -c w + c B · '· (4) Since the present invention uses the relative method, the scattering coefficient s w is independent of the mixing ratio c w of the white pigment.
= 1 と仮定する。 また、 通常はベース釉の調合率 CB を一定とし、 白色顔料の調 合率 cwを変化させて混合物を作成するので、 CB =〗 0 0 (—定) と仮定する。 この時、 白色顔料の調合率 „ は、 いわゆる外添加調合率 (ベ一ス釉の重量を 1 0 0 %とした時の顔料の重量%) となる。 これらの仮定をおくと、 上記数式 4は 次の数式 5のように害き換えられる。 Cw+ 100SB = 1 Also, usually, the blending ratio C B of the base glaze is constant and the blending ratio c w of the white pigment is changed to form a mixture. Therefore, it is assumed that C B =〗 0 0 (—constant). At this time, the mixing ratio の of the white pigment is the so-called external addition mixing ratio (weight% of the pigment when the weight of the base glaze is 100%). Is harmed by the following equation 5. C w + 100S B
Sw'= Cw+ 100 · ' · ( 5 ) ベース釉は無色透明に近いが、 完全な無色透明ではない。 すなわち、 数式 5に おいて、 ベ一ス釉の散乱係数 SB は 0でない正の小さな値である。 従って、 ベー ス釉と白色顔料の混合物の散乱係数 Sw' は、 図 2 (A) に示すようなグラフと なる。 このグラフは、 白色顔料の調合率 Cw が大きくなると Sw' =〗に漸近す る双曲線である。 Sw ' = Cw + 100 ·' · (5) The base glaze is almost colorless and transparent, but not completely colorless and transparent. That is, in Equation 5, the scattering coefficient S B of the base glaze is a small positive non-zero value. Therefore, the scattering coefficient S w 'of the mixture of the base glaze and the white pigment becomes a graph as shown in Fig. 2 (A). This graph is a hyperbola that gradually approaches S w ′ =〗 as the blending ratio C w of the white pigment increases.
ところで、 図 2 (A) に示すような散乱係数 Sw' を実測することは容易では ない。 そこで、 この第 1実施例では、 白色顔料の調合率 Cwに対する散乱係数 Sw— , のグラフの形状を、 図 2 (B) に示すように高濃度側の第 1の直線 L 1と低濃 度側の第 2の直 L 2で近似できるものと仮定した。 直線し L 2を表わす関数 S = f , (Cw ) , S = f 2 (Cw ) は、 それぞれ次の数式 6および 7で表 わすことができる。 Sw' = aCw + b -..(6) (C„。≤ Cw) By the way, it is not easy to measure the scattering coefficient S w ′ as shown in Fig. 2 (A). Therefore, in the first embodiment, the shape of the graph of the scattering coefficient S w —, with respect to the mixing ratio C w of the white pigment is changed to the first straight line L 1 on the high concentration side as shown in FIG. It is assumed that it can be approximated by the second straight line L 2 on the concentration side. The functions S = f, (C w ) and S = f 2 (C w ), which are linear and represent L 2, can be expressed by the following equations 6 and 7, respectively. S w '= aC w + b-.. (6) (C „.≤ C w )
Sw' = d Cw + e ... (7) (CK < Cwo) S w '= d C w + e ... (7) (C K <C wo )
ここで, は直線 L I, L 2の交点における白色顔料の調合率である。 Here, is the blending ratio of the white pigment at the intersection of the straight lines L I and L 2.
数式 6, 7における係数 a, b, d, eを決定すれば、 (ベ一ス釉 +白色顔料) の混合物に対する散乱係数 Sw, が求まる。 そして、 散乱係数 Sw, が求まれば、 分光反射率 R (A ) の測定値を用いて上記数式 2から吸収係数 Kw' を求めるこ とができる。 By determining the coefficients a, b, d, and e in Equations 6 and 7, the scattering coefficient S w , for the mixture of (base glaze + white pigment) is determined. Then, if the scattering coefficient S w , is obtained, the absorption coefficient K w ′ can be obtained from Equation 2 using the measured value of the spectral reflectance R (A).
図 3は、 図 1のステップ S 1の詳細手順を示すフローチャートである。 ステツ プ S 1 〗では、 ベース釉と白色顔料のみを混合した混合物 (第 1の混合物) を作 成した。 図 4は、 第〗実施例において準備した 1 3個のサンプル W】 2〜W0の 調合率を示している。 図 4から解るように、 1 3個のサンプル Wl 2〜W0は、 白色顔料の調合率 Cw を 1 2%〜0%の範囲で 1 %ずつ変化させて作成したもの である。 なお、 第 1実施例におけるサンプルは、 通常の陶磁器の素地に、 調合し た釉をかけて焼成することによって作成した。 また、 後述する他のサンプルも同 じ条件で作成した。 FIG. 3 is a flowchart showing a detailed procedure of step S1 in FIG. In step S 1。, a mixture (first mixture) was prepared in which only the base glaze and the white pigment were mixed. FIG. 4 shows the mixing ratios of 13 samples W] 2 to W0 prepared in the first embodiment. As can be seen from FIG. 4, 1 3 samples Wl 2~W0 are those created by the compounding ratio C w of the white pigment is changed by 1% in the range of 1 2% 0%. The sample in the first example was prepared by baking a normal ceramic body with a blended glaze. In addition, other samples described later were prepared under the same conditions.
ステップ S 1 2では、 各サンプル W〗 2〜W0について、 分光光度計で分光反 射率 R' (A ) を測定した。 ステップ S 1 3では、 分光反射率の測定値 R, In step S12, the spectral reflectance R '(A) of each sample W2 to W0 was measured with a spectrophotometer. In step S13, the measured values of spectral reflectance R,
(λ ) に基づいて、 数式 6に舍まれる係数 a, bを以下のように実験的に決定し た。 Based on (λ), the coefficients a and b in Equation 6 were experimentally determined as follows.
白色顔料の調合率 Cw が比較的大きなサンプルでは、 釉の隠ぺぃ力が大きいの で、 Sw' = 1と仮定することができる。 具体的には、 白色顔料の調合率 Cw が 最も大きなサンプル W1 2に対する散乱係数 Sw' を 1に等しいと仮定した。 こ の仮定を用いると、 白色顔料の調合率 Cw が i %のサンプル W iの散乱係数 Swi_ ' は、 次の数式 8で与えられる。 In a sample having a relatively large blending ratio C w of the white pigment, the hiding power of the glaze is large, so it can be assumed that S w ′ = 1. Specifically, it was assumed that the scattering coefficient S w ′ for the sample W 12 having the largest white pigment preparation ratio C w was equal to 1. Using this assumption, the scattering coefficient S wi ′ of the sample W i having the white pigment preparation rate C w of i% is given by the following equation (8).
。w i = Swn ' ...(8). w i = Swn '... (8)
Figure imgf000019_0001
CT/JP96/00738
Figure imgf000019_0001
CT / JP96 / 00738
18 数式 8において、 (KZS) W12 は白色顔料の調合率が 1 2 %のサンプル W 1 2の吸収係数と散乱係数の比であり、 (KZS) wiは白色顔料の調合率が i %の サンプル W iの吸収係数と散乱係数の比である。 また、 数式 8では、 吸収係数の 変化は散乱係数の変化に比べて小さく、 Kwl 2 ==Kw iであると仮定している。 数式 8の右辺の (KZS) wl2 , (K/S) wiはサンプル W iの分光反射率の 測定 :R' ( J から、 上述した数式 2に従って求めることができる。 似し、 混 合物である釉の厚みは小さいので、 分光光度計で測定される分光反射率 R, (λ ) を、 次の数式 9 (サンダーソン (Saunderson) の式) に従って理想状態18 In Equation 8, (KZS) W12 is the ratio of the absorption coefficient to the scattering coefficient of the sample W12 in which the blending ratio of the white pigment is 12 %, and (KZS) wi is the sample in which the blending ratio of the white pigment is i%. It is the ratio between the absorption coefficient and the scattering coefficient of Wi. Equation 8 assumes that the change in the absorption coefficient is smaller than the change in the scattering coefficient, and that K wl 2 == K wi . (KZS) wl2 , (K / S) wi on the right side of Equation 8 can be obtained from the measurement of the spectral reflectance of the sample W i: R ′ (J according to Equation 2 described above. Because the thickness of a certain glaze is small, the spectral reflectance R, (λ) measured by a spectrophotometer is calculated according to the following equation 9 (Saunderson's equation).
(混合物の厚みが無限大の状態) の分光反射率 R (; - ) に変換する。 p - k - k】 Q (The thickness of the mixture is infinite) and is converted into the spectral reflectance R (;-). p-k-k] Q
(1 -ki)(l -k2) + k2R'- kjk2 · ここで、 係数 , k 2 は、 被着色物 (ベース釉) の光学的性質に依存する 値である。 係数 , k 2 としては、 被着色物の屈折率 nから次の数式 1 0に 従って決定することができる。
Figure imgf000020_0001
(1 -ki) (l -k 2 ) + k 2 R'-kjk 2 · Here, the coefficient, k 2 is a value that depends on the optical properties of the object to be colored (base glaze). The coefficients, k 2 can be determined from the refractive index n of the object to be colored according to the following equation 10:
Figure imgf000020_0001
k2=0.68n-0.56 ... (10b) k 2 = 0.68n-0.56 ... (10b)
なお、 第 1実施例において用いたベース釉は、 屈折率 πが約 1. 4である。 こ うして理想状態の分光反射率 R (λ ) が求まると、 上述の数式 2から (KZS) wiが得られる。 The base glaze used in the first embodiment has a refractive index π of about 1.4. When the spectral reflectance R (λ) in the ideal state is obtained in this way, (KZS) wi is obtained from the above-described equation (2).
数式 9に従って分光反射率の測定値 R' U ) から理想状態の分光反射率 R {1 ) を求める演算は、 後述する他の工程においてサンプルの分光反射率から (K/S) を求める際にも同様に実施される。  The calculation to determine the ideal state spectral reflectance R (1) from the spectral reflectance measured value R 'U) according to Equation 9 is performed when calculating (K / S) from the spectral reflectance of the sample in another process described later. Is similarly performed.
各サンプル W iについて得られた (K/S) wiの値を上記数式 8に代入するこ とによって、 各調合率 Cw において混合物の散乱係数 S„' の値を求めることが でき、 係数 a, bを決定することができる。 係数 a, bを含む数式 6は、 白色顔 料の調合率 C„ が比較的髙ぃ範囲にのみ適用される。 従って、 係数 a, bの決定 においても、 調合率 Cw が比較的高い範囲の (例えば 1 0%〜 1 2%の) 数個の サンプルの (K/S) wiのみを用いる。 The obtained values of (K / S) wi by the child substituted into the equation 8 for each sample W i, it is possible to determine the value of the scattering coefficient S "'of the mixture in each of the mixing ratio C w, the coefficient a , B can be determined Equation 6 including the coefficients a and b is applied only to the range where the mixing ratio C „of the white pigment is relatively high. Therefore, determination of coefficients a and b In (2), only (K / S) wi of several samples in which the blending ratio Cw is relatively high (for example, 10% to 12%) is used.
図 5は、 Cw= l 0%〜〗 2%のサンプルの分光反射率 R U ) を示す概念図 である。 このように、 ベース釉に高濃度の白色顔料を混合したものは、 可視光線 の波長 の全範囲 (約 400 n m〜約 700 n m) においてほとんど一定の反射 率を示す。 そこで、 この分光反射率 R (λ ) の値から数式 2と数式 8を用いて散 乱係数 Swi' を求めることができる。 なお、 この散乱係数 S は、 可視光線の 全波長範两において同一の値を有する。 第 1実施例では、 a = 0. 0005, b =0. 94が得られたので、 数式 6が次の数式 1 1に書き換えられた。 Figure 5 is a conceptual diagram showing a spectral reflectance RU) of C w = l 0% ~〗 2% sample. Thus, the base glaze mixed with a high concentration of white pigment exhibits almost constant reflectance over the entire visible wavelength range (about 400 nm to about 700 nm). Then, the scattering coefficient S wi ′ can be obtained from the value of the spectral reflectance R (λ) using Expressions 2 and 8. The scattering coefficient S has the same value in the entire wavelength range of visible light. In the first embodiment, since a = 0.0005 and b = 0.94 were obtained, Expression 6 was rewritten into Expression 11 below.
Sw ' = 0.005 Cw + 0. 4 ...(11) (髙濃度側) S w '= 0.005 C w + 0.4 ... (11) (髙 concentration side)
なお、 数式 1 1は、 Cw = 1 0%付近において Sw' = 1に極めて近い値を与 えるので、 数式 1 1の代わりに Sw, = 1を用いてもよい。 Since Equation 11 gives a value very close to S w ′ = 1 near C w = 10%, S w , = 1 may be used instead of Equation 11.
図 3のステップ S 1 4では、 低濃度側の数式 7の係数 d, eの値を決定する。 白色顔料の調合率 Cw が比較的小さい場合には下地の影響が大きくなるので、 分 光反射率 R (λ ) の測定値が真の釉の層のみの情報ではなくなる。 従って、 分光 反射率 R (λ ) の測定値に基づいて係数 d, eを決定するのは容易ではない。 そ こで、 係数 d, eの値は、 図 6に示すような手順に従って決定する。 In step S14 of FIG. 3, the values of the coefficients d and e of Equation 7 on the low concentration side are determined. When the blending ratio C w of the white pigment is relatively small, the influence of the ground becomes large, so that the measured value of the spectroscopic reflectance R (λ) is not information of only the true glaze layer. Therefore, it is not easy to determine the coefficients d and e based on the measured values of the spectral reflectance R (λ). Therefore, the values of the coefficients d and e are determined according to the procedure shown in Fig. 6.
ステップ S 2 1では、 顔料物性値決定用サンプルを作成する。 図 7は、 顔料決 定用サンプルの調合率を示す説明図である。 顔料物性値決定用サンプルは、 ベー ス袖と白色顔料と他の有色顔料を混合して作成したものである。 第 1実施例では、 白色顔料と有色顔料の合計の調合率 (Pigment Volume Concentration, P VC) を 1 2%—定とし、 有色顔料の調合率 Cp を 1 %〜 1 2%の範囲で 1 %ず つ変えた 1 2個のサンプル M】〜M 1 2を作成した。 なお、 有色顔料としては、 陶磁器の彩色に良く用いられる青色、 赤色、 黄色、 等を選択し、 各顔料について 図 7に示すような 1 2個のサンプル M〗〜M 1 2をそれぞれ作成した。 In step S21, a sample for determining the physical properties of a pigment is prepared. FIG. 7 is an explanatory diagram showing the mixing ratio of the pigment determination sample. The sample for determining the physical properties of the pigment was prepared by mixing a base sleeve, a white pigment and another colored pigment. In the first embodiment, the total compounding ratio of the white pigment and colored pigment (Pigment Volume Concentration, P VC) 1 2% - a constant, 1 formulatory ratio C p of colored pigment in the range of 1% to 1 2% 12 samples M] to M12 were prepared. As colored pigments, blue, red, yellow, etc., which are often used for coloring ceramics, were selected, and for each pigment, 12 samples M〗 to M12 as shown in FIG. 7 were prepared.
なお、 サンプル M l 2は白色顔料を舍まない混合物であり、 本願発明の第 2の 混合物に相当する。 また、 サンプル M 1〜M 1 1は本願発明の第 3の混合物に相 当する。 Note that the sample Ml2 is a mixture containing no white pigment, and corresponds to the second mixture of the present invention. Samples M1 to M11 correspond to the third mixture of the present invention. Hit.
ステップ S 2 2では、 検証用サンプルを作成する。 |、|8は、 第 1実施例で用い た検証用サンプルの調合率を示す説明図である。 4つの検証用サンプル D 1〜D 4は、 4色の顔料をベース釉に混合したものである。 なお、 ベース釉の調合率 C B は省略しているが、 いずれのサンプルも CB = 1 00%である。 In step S22, a verification sample is created. |, | 8 are explanatory diagrams showing the mixing ratios of the verification samples used in the first example. The four verification samples D1 to D4 are four-color pigments mixed in the base glaze. Incidentally, Formulation ratio C B of the base glaze is omitted, all of the samples are also C B = 1 00%.
検証用サンプル D〗〜D 4は、 本願発明における第 4の混合物に相当する。 ステップ S 2 3では、 係数 d, eに仮の値を割り当てる。 図 2 (A) , (B) を比校すれば解るように、 係数 eの値は、 ベース釉の散乱係数 SB の値に相当し ており、 かなり小さな値であることが解っている。 例えば、 e = 0と仮定するこ とも可能である。 また、 図 2 (B) に示す第 1 と第 2の直線 L 1, L 2の交点は、 横轴の調合率 Cw が数%のところに位置するので、 係数 dも概略の値を仮決めす ることが可能である。 The verification samples D〗 to D4 correspond to the fourth mixture in the present invention. In step S23, temporary values are assigned to the coefficients d and e. As can be seen by comparing Figs. 2 (A) and 2 (B), the value of the coefficient e corresponds to the value of the scattering coefficient S B of the base glaze, and is a very small value. For example, it is possible to assume e = 0. Also, since the intersection of the first and second straight lines L 1 and L 2 shown in FIG. 2 (B) is located at a position where the horizontal mixing ratio C w is several percent, the coefficient d is also a rough value. It is possible to decide.
係数 d, eを仮決めすると、 調合率 Cw に依存する散乱係数 Swi' が調合率 Cw の全範囲において決定される。 散乱係数 SWi' が決まると、 次の数式 1 2に従つ て調合率 Cw に依存する吸収係数 Kwi' も得られる。 Wi ' = ( ■) xSw,' ...(12) Coefficient d, the temporarily decided e, scattering coefficient depends on the compounding ratio C w S wi 'is determined in the entire range of the compounding ratio C w. When the scattering coefficient S Wi 'is determined, an absorption coefficient K wi ' depending on the mixing ratio C w is obtained according to the following equation (12). W i '= (■) xSw,' ... (12)
、 へ; なお、 散乱係数 Swi' の値は波長スに依存しないが、 吸収係数 Kw,' の値は波 長 に依存する。 この理由は、 この発明の相対法においては散乱係数 swi' を波 長スに依存しない基準値として用い、 他の物性値の相対値を求めるからである。 分光反射率 R U) の値はもちろん波長スに依存するので、 数式 2に従って得ら れる (KZS) wiの愤も波長スに依存する。 従って、 数式 1 2に従って得られる 吸収係数 Kwi' も波長スに依存する。 換言すれば、 吸収係数 Kwi' は白色顔料の 調合率 Cw と波長とに依存する形式で求められる。 The value of the scattering coefficient S wi 'does not depend on the wavelength, but the value of the absorption coefficient K w ,' depends on the wavelength. The reason for this is that in the relative method of the present invention, the scattering coefficient s wi ′ is used as a reference value that does not depend on the wavelength, and the relative values of other physical properties are obtained. Since the value of the spectral reflectance RU) depends of course on the wavelength, the value of (KZS) wi obtained according to Equation 2 also depends on the wavelength. Therefore, the absorption coefficient K wi 'obtained according to Equation 12 also depends on the wavelength. In other words, the absorption coefficient K wi 'is obtained in a form that depends on the blending ratio C w and the wavelength of the white pigment.
ステップ S 25では、 顔料物性値決定用サンプルを用いて、 各有色顔料の吸収 係数 ΚΡ (λ ) と散乱係数 S p (λ ) とを求める。 なお、 ステップ S 25は、 図 1におけるステップ S 2に相当する。 ( . ) と S P (ス) は、 次のような手 順で算出される。 まず、 顔料物性値決定用サンプルについては、 ヒ述の数式 1から次の数式 1 3 が得られる。 In step S25, the absorption coefficient Κ Ρ (λ) and the scattering coefficient S p (λ) of each colored pigment are determined using the pigment property value determination sample. Step S25 corresponds to step S2 in FIG. (.) And SP (s) are calculated by the following procedure. First, for a pigment physical property value determination sample, the following expression 13 is obtained from expression 1 described above.
(CR + ρ)Κ\ι _ ^ΒΚΒ + v^ K\v + Cprvp (1 (CR + ρ) Κ \ ι _ ^ Β Κ Β + v ^ K \ v + Cprvp (1
(C B + C\v + Cp)a.\i CB¾ W W + Cp ところで、 上記の数式 3は、 次の数式 1 4のように害き換えられるので、 数式 1 4を数式〗 3に代入すると、 数式 1 5が得られる。  (CB + C \ v + Cp) a. \ I CB¾ WW + Cp By the way, the above formula 3 can be replaced by the following formula 14.So, when formula 14 is substituted into formula〗 3, Equation 15 is obtained.
Cw T V ' = CB¾ + Cw \v · · · 4a )  Cw T V '= CB¾ + Cw \ v 4a)
Cw 'S w 1 = CBSR + CWSW . - - (14b) C w 'S w 1 = C B S R + C W S W .--(14b)
Cw * = CB + Cw ... (14c) Cw * = C B + C w ... (14c)
KM _ C\y ' Kv ' + C K · ( ) KM _ C \ y 'Kv' + C K · ()
SM C\y ' Sy ' + CpSp 数式〗 5を変形すると、 次の数式 1 6が得られる。  SM C \ y 'Sy' + CpSp By transforming equation〗 5, the following equation 16 is obtained.
Figure imgf000023_0001
ここで、 (KZS) p は有色顔料単独での吸収係数と散乱係数の比であり、 (K/S) は顔料物性値決定用サンプルの吸収係数と散乱係数の比、 (KZ S) は (ベース釉 +白色顔料) のサンプルの吸収係数と散乱係数の比である。 また、 cw, はベース釉と白色顔料の調合率の合計値、 s は (ベース釉 +白色 顔料) の混合物の散乱係数であり、 数式 6および 7で与えられる値である。 数式 1 6の右辺の各項は、 以下のように求めることができる。
Figure imgf000023_0001
Here, (KZS) p is the ratio between the absorption coefficient and the scattering coefficient of the colored pigment alone, (K / S) is the ratio between the absorption coefficient and the scattering coefficient of the pigment physical property determination sample, and (KZ S) is ( This is the ratio between the absorption coefficient and the scattering coefficient of the sample (base glaze + white pigment). C w , is the sum of the blending ratios of the base glaze and the white pigment, and s is the scattering coefficient of the mixture of (base glaze + white pigment), which is given by Equations 6 and 7. The terms on the right side of Equation 16 can be obtained as follows.
(K/S) p の倘は、 白色顔料を舍まない顔料物性値決定用サンプル (図 7の サンプル M l 2) の分光反射率 R (λ ) から上記数式 2によって求めることがで きる。 但し、 厳密に言えば、 数式 1 6における (KZS) の値はベース釉の影 響を含まない有色顔料^独に対する値である。 一方、 上述のようにして実測で得 られる (KZS) ρ は (ベース釉 +有色顔料) のサンプルに対する他なので、 ベー ス釉の影響が舍まれている。 し力 し、 17のサンプル M 1 2の顔料の調合率は 1 2%と卨ぃ値なので、 (KZS) p に対するベース釉の寄与は極めて小さい。 従って、 数式 1 6における (K/S) p の値としては、 サンプル M 1 2の分光反 射率 R (λ ) から数式 2によって求めた値を用いても、 その誤差は無視できる程 度である。 K of (K / S) p can be obtained by the above equation 2 from the spectral reflectance R (λ) of the pigment physical property determination sample (sample Ml 2 in FIG. 7) which does not contain white pigment. However, strictly speaking, the value of (KZS) in Equation 16 is a value for a colored pigment that does not include the effect of the base glaze. On the other hand, (KZS) ρ obtained by actual measurement as described above is for the sample of (base glaze + colored pigment). The influence of the glaze is impaired. However, since the mixing ratio of the pigment of 17 samples M12 is 12%, which is a 卨 ぃ value, the contribution of the base glaze to (KZS) p is extremely small. Therefore, as the value of (K / S) p in Equation 16, even if the value obtained from Equation 2 from the spectral reflectance R (λ) of sample M 12 is used, the error is negligible. is there.
(KZS) Μ の値は、 白色顔料と有色顔料の両方を舍む顔料物性値決定用サン プル (図 7のサンプル Μ 1〜! VI 1 1 ) の分光反射率 R (;. ) から数式 2に従って 算出できる。 従って、 (KZS) Μ は有色顔料の調合率 Cp に依存した形式で求 められる。 (KZS) value of Micromax is the spectral reflectance of the white pigment and Complex both colored pigments unpigmented physical properties determined for samples (samples of Figure 7 Μ 1~ VI 1 1!) R (;.) From the formula 2 Can be calculated according to Therefore, (KZS) Μ is determined in a form depending on the mixing ratio C p of the colored pigment.
(K/S) w' の値は、 4に示す (ベース釉 +白色顔料) のサンプルの分光 反射率 R (ス) 力 ^数式 2に従って算出された値を用いることができる。 As the value of (K / S) w ′, the value calculated according to the spectral reflectance R (s) power of the sample of (base glaze + white pigment) shown in 4 ^ Equation 2 can be used.
以上のようにして数式 1 6の右辺の各項の値を求めることができるので、 顔料 物性値決定用サンプル Ml〜M 1 1のそれぞれについて、 散乱係数 Sp (λ ) を 求めることができる。 なお、 散乱係数 Sp (Λ) は、 有色顔料の調合率 Cp に依 存する形式で求められる。 It is possible to determine the value of each term of the to the right side of Equation 1 6 as described above, for each of the pigment physical properties determined for the samples Ml~M 1 1, can be obtained scattering coefficient S p a (lambda). The scattering coefficient S p (Λ) is obtained in a form depending on the mixing ratio C p of the colored pigment.
こうして得られた散乱係数 Sp (λ ) と、 分光反射率 R ( .) の測定値から数 式 2で算出された (KZS) ρ の値とに基づいて、 次の数式 1 7に従って吸収係 数 ΚΡ {?. ) も求められる。 Based on the scattering coefficient S p (λ) thus obtained and the value of (KZS) ρ calculated from the measured value of the spectral reflectance R (. The number Κ Ρ {?.) Is also required.
(17) (17)
. なお、 吸収係数 Κρ (λ ) は、 散乱係数 S p (λ ) と同様に、 有色顔料の調合 率 Cp に依存する形式で求められる。 . The absorption coefficient Κ ρ (λ), as well as the scattering coefficient S p (λ), obtained in a form which depends on the compounding ratio C p of colored pigments.
図 9は、 吸収係数 ΚΡ (Λ) の調合率 Cp に対する依存性の一例を示すグラフ である。 散乱係数 S p (; J もこれと同様な依存性を示す。 なお、 図 9では、 代 表的な波長 についてのグラフのみを示しているが、 実際には可視光線の全波長 範囲 (約 400 nm〜約 7 00 n m) で〗 0 n m毎に吸収係数 Κ ρ (λ ) が求め られる。 以上のようにして、 (ベース釉 +白色顔料) の混合物の吸収係数 Kw, (ス) および散乱係数 Sw' と、 他の有色顔料の吸収係数 KP ) および散乱係数 S p (; J が求まると、 図 6のステップ S 26において、 検証用サンプル (図 8) に 対してコンピュータカラーマッチングによるシミュレーションを行なう。 FIG. 9 is a graph showing an example of the dependence of the absorption coefficient Κ Ρ (Λ) on the mixing ratio C p . The scattering coefficient S p (; J shows a similar dependence. Note that FIG. 9 shows only a graph of typical wavelengths, but in actuality, the actual wavelength range of visible light (about 400 nm~ about 7 00 nm) absorption coefficient for each〗 0 nm in Κ ρ (λ) is obtained. As described above, the absorption coefficient K w , (s) and the scattering coefficient S w ′ of the mixture of (base glaze + white pigment) and the absorption coefficient K p of other colored pigments and the scattering coefficient S p (; J When is obtained, in step S26 in FIG. 6, a simulation is performed on the verification sample (FIG. 8) by computer color matching.
ここで、 コンピュータカラーマッチングによる色予測と調合割合の予測につい て簡単に説明する。 数式 2を変形すると、 混合物の分光反射率 R ) は、 次の 数式 1 8で与えられる。  Here, the color prediction by computer color matching and the prediction of the mixing ratio will be briefly described. By transforming Equation 2, the spectral reflectance R) of the mixture is given by Equation 18 below.
(18) (18)
' MM
Figure imgf000025_0001
混合物の吸収係数と散乱係数の比 (KZS) M は、 数式 1に従って算出できる ので、 混合物の分光反射率 R (λ ) も数式 1 8から求めることができる。 この分 光反射率 R (λ ) は理想状態の分光反射率なので、 分光光度計で測定できる分光 反射率 (λ) を、 数式 9 (サンダーソンの式) を変形した次 式 19に従つ て求める。
'' MM
Figure imgf000025_0001
Since the ratio (KZS) M of the absorption coefficient and the scattering coefficient of the mixture can be calculated according to Equation 1, the spectral reflectance R (λ) of the mixture can also be determined from Equation 18. Since this spectroscopic reflectance R (λ) is an ideal state spectral reflectance, the spectral reflectance (λ) that can be measured by a spectrophotometer is calculated according to the following Equation 19, which is a modification of Equation 9 (Sanderson's equation). Ask.
R' = ki + (l-ki)(l-k2)] ...(19) 係数 , k2 は上記数式〗 0で与えられる。 R ′ = ki + (l-ki) (lk 2 ) ] ... (19) The coefficient, k 2 is given by the above equation〗 0.
分光反射率 R' (λ ) が求まると、 混合物の三刺激値 X, Υ, Ζが次の数式 2 0によって求められる。  Once the spectral reflectance R '(λ) has been determined, the tristimulus values X, Υ, Ζ of the mixture are determined by the following equation 20.
• .. (20a) • .. (20a)
Y = k(S( )R )yWd .. - (20b) Z=kjS( )R'( )z( )dX ••■(20c) k =, 100 ...(20d) Y = k (S () R) yWd ..-(20b) Z = kjS () R '() z () dX •• ■ (20c) k =, 100 ... (20d)
ここで、 S ) は標準光の分光分布、 X (え) , y (λ ) , ζ (λ ) (数式 中ではバ一付きである) は等色関数である。 コンピュータカラーマッチングによって混合物の色を予測する場合には、 上述 のように、 混合物の各成分の吸収係数と散乱係数に基づき、 数式 数式 1 8〜 20に従って、 その混合物の三刺激値 X, Υ, Zを算出する。 三刺激値は混合物 の色を表わすので、 任意の混合物の色を予測することができることになる。 また、 所望の色を有する混合物の調合割合を予測する場合には、 混合物の調合 割合を仮定して、 上述した手順でその三刺激値を算出し、 ニュートン一ラプソン 法などの逐次近似法によって所望の色に所定の誤差内で一致するような調合割合 を求める。 Here, S) is the spectral distribution of the standard light, and X (e), y (λ), ζ (λ) (in the formulas, with a dash) are color matching functions. When predicting the color of a mixture by computer color matching, as described above, based on the absorption coefficient and the scattering coefficient of each component of the mixture, the tristimulus values X, Υ, Calculate Z. Since the tristimulus values represent the color of the mixture, the color of any mixture can be predicted. Further, when predicting the blending ratio of a mixture having a desired color, the tristimulus value is calculated by the above-described procedure assuming the blending ratio of the mixture, and the desired value is calculated by a successive approximation method such as the Newton-Raphson method. Find a blending ratio that matches the color of the product within a specified error.
なお、 丫∑表色系の代ゎりにし* 3*1)* 表色系のような他の表色系を用い てコンピュータカラ一マッチングを行なうことも可能である。  Note that it is also possible to perform computer color matching using another color system such as * 3 * 1) * color system instead of the color system.
図 6のステップ S 26におけるシミュレーションでは、 コンピュータカラーマツ チングによって、 図 8に示す各検証用サンプル D 1〜D 4の調合率を予測する。 検証用サンプルの調合率は既知なので、 予測した調合率と実調合率との一致度は 簡単に算出できる。 例えば、 調合率の一致度の指標として、 次の数式 2 1で与え られる自乗平均誤差厶を使用する。  In the simulation in step S26 in FIG. 6, the compounding ratio of each of the verification samples D1 to D4 shown in FIG. 8 is predicted by computer color matching. Since the mixing ratio of the verification sample is known, the degree of coincidence between the predicted mixing ratio and the actual mixing ratio can be easily calculated. For example, the root mean square error given by the following equation 21 is used as an index of the degree of coincidence of the mixing ratio.
> (CR.-C,)2 > (CR.-C,) 2
Δ=-^ · '·(21) Δ = - ^ · '· ( 21)
ここで は成分 iの実調合率、 Ci は予測された調合率である。 ステップ S 26において、 各検証用サンプル D 1〜D 4について得られた自乗平均誤差八が 所定の許容値以下でない場合にはステップ S 2 3に戻り、 係数 d, eの値を修正 する。 こうして、 各検証用サンプル D 1〜D 4について得られる自乗平均誤差△ が所定の許容値以下になるまでステップ S 23〜S 26を繰り返し実行し、 この 結果、 係数 d, eを決定することができる (ステップ S 2 7) 。 第 1実施例では、 d = 0. 1 225, e = 0となり、 数式 7が次の数式 22書き換えられた。  Where is the actual blending ratio of component i and Ci is the predicted blending ratio. In step S26, if the root mean square error 8 obtained for each of the verification samples D1 to D4 is not equal to or smaller than the predetermined allowable value, the process returns to step S23 to correct the values of the coefficients d and e. Thus, steps S23 to S26 are repeatedly executed until the root mean square error に つ い て obtained for each of the verification samples D1 to D4 is equal to or smaller than a predetermined allowable value. As a result, the coefficients d and e can be determined. Yes (step S27). In the first embodiment, d = 0.125, e = 0, and Expression 7 is rewritten as Expression 22 below.
Sw ' = 0.1225 Cw . - . (22) (低浪度側) S w '= 0.1225 C w - .. (22) ( low Sina degree side)
なお、 数式 1 〗で与えられる高渙度側の散乱係数 Sw' のグラフ (図 2 (B) の直線 L I) と、 数式 22で与えられる低濃度側の散乱係数 Sw' のグラフ (直 線 L 2) との交点における調合率 Cw は 8. 0%である。 Note that the graph of the scattering coefficient S w 'on the high lysis level given by Equation 1 (Fig. 2 (B) The mixing ratio C w at the intersection of the straight line LI) of this and the graph of the scattering coefficient S w ′ on the low concentration side given by Equation 22 (the straight line L 2) is 8.0%.
以上のように、 上記第 1実施例では検証用サンプルの調合率を精度良く予測で きるように数式 7における係数 d, eを決定したので、 A色顔料の調合率 Cw が 比較的低い領域においても、 (ベース釉 +白色顔料) の混合物の散乱係数 Sw' と吸収係数 Kw' (λ) とを精度良く決定することができた。 As described above, coefficients in Equation 7 as cut with accurately predicting the formulation ratio of the verification samples in the first embodiment d, since the determined e, Formulation ratio C w of the A color pigment is relatively low area In, the scattering coefficient S w 'and the absorption coefficient K w ' (λ) of the mixture of (base glaze + white pigment) could be determined with high accuracy.
なお、 図 6に示す手順で (ベ一ス釉 +白色顔料) の混合物の散乱係数 Sw' と 吸収係数 K (/) とが求められる際には、 ステップ S 25において他の有色 顔料の吸収係数 ΚΡ (λ ) と散乱係数 Sp (λ ) も同時に求められる。 When the scattering coefficient S w ′ and the absorption coefficient K (/) of the mixture of (base glaze + white pigment) are determined by the procedure shown in FIG. 6, the absorption of other colored pigments is determined in step S25. coefficient Κ Ρ (λ) and scattering coefficient S p (λ) is also required at the same time.
こうして各成分の吸収係数と散乱係数が求まると、 図 1のステップ S 3におい て、 コンピュータ力ラーマッチングによる色予測や調合割合の予測が実行される。 図 1 0は、 第 1実施例における調合率の予測結果を示す説明図である。 図 1 0に おける第 1実施例の予測結果は、 数式 1 1および数式 22を用いた場合の結果で ある。 また、 比較例の予測結果は、 S„' = 0 (一定) の仮定 (図 2 (B) において破線で示すグラフ) を用いた場合の結果である。 なお、 図 1 0の第 1実 施例と比較例では、 L*a*b* 表色系を用いて色差が最小となる調合率を予測 した。  When the absorption coefficient and the scattering coefficient of each component are obtained in this way, in step S3 in FIG. 1, color prediction and blend ratio prediction by computer color matching are executed. FIG. 10 is an explanatory diagram showing a prediction result of the mixing ratio in the first embodiment. The prediction result of the first embodiment in FIG. 10 is a result obtained when Expressions 11 and 22 are used. In addition, the prediction result of the comparative example is a result when the assumption of S „′ = 0 (constant) (a graph shown by a broken line in FIG. 2B) is used. In the examples and comparative examples, the blending ratio that minimizes the color difference was predicted using the L * a * b * color system.
図 1 0に示されるように、 第]実施例による予測結果は調合率の自乗平均誤差 が比較例に比べてかなり小さく、 より高精度で調合率を予測できたことが解る。 特に、 サンプル P 2, P 3のように白色顔料の調合率が小さな場合には、 比較例 の予測精度がかなり悪化しているのに対して、 第 1実施例の予測精度は良好であ る。 この理由は、 図 2 (A) , (B) に示すように、 白色顔料の調合率 Cw が小 さくなるに従って、 実際の散乱係数 Sw, の値が 1から離れていくので、 Sw, = 1 (一定) と仮定した比較例では、 各成分の吸収係数と散乱係数の値の誤差が大 きくなるからであると考えられる。 As shown in FIG. 10, the prediction result of the Example] has a significantly smaller root mean square error of the mixing ratio than that of the comparative example, indicating that the mixing ratio can be predicted with higher accuracy. In particular, when the mixing ratio of the white pigment is small as in the samples P2 and P3, the prediction accuracy of the comparative example is considerably deteriorated, whereas the prediction accuracy of the first embodiment is good. . This is because, as shown in FIG. 2 (A), (B) , in accordance with Formulation ratio C w of the white pigment is small fence, the actual scattering coefficient S w, the value of moves away from 1, S w , = 1 (constant), it is considered that the error between the absorption coefficient and the scattering coefficient of each component is large.
以上のように、 この第 1実施例では、 無色透明でない被着色物であるべ一ス釉 の光学的影響を考慮して各成分の吸収係数と散乱係数とを決定しているので、 コ ンピュータカラ一マッチングによつて高精度の予測を行なうことが可能である。 また、 サンプルとしては図 4、 図 7、 図 8に示すような混合物を通常の素地の上 に塗布したものを準備すればよいので、 サンプルの作成も容易である。 As described above, in the first embodiment, since the absorption coefficient and the scattering coefficient of each component are determined in consideration of the optical effect of the base glaze, which is not colorless and transparent, the computer is used. High-precision prediction can be performed by color matching. In addition, it is only necessary to prepare a sample as shown in Figs. 4, 7 and 8 on which a mixture is applied on a normal substrate, so that it is easy to prepare a sample.
なお、 上記第 1実施例に関しては、 例えば次のような変形も可能である。  Note that, for the first embodiment, for example, the following modifications are possible.
( 1) 散乱係数 Sw' の値を図 2 (B) に示される 2本の直線 L 1, L 2を数式 6, 7で近似する代わりに、 数式 5をそのまま用いて散乱係数 Sw' を決定して もよい。 この場合には、 図 6のステップ S 2 3においてベース釉の散乱係数 SB の値を仮決定し、 ステップ S 2 3〜S 26を繰り返し実行することによって散乱 係数 SB の値を求めるようにすればよい。 こうすれば、 未知数が散乱係数 SB(1) scattering coefficient S w 'linear L 1 2 pieces of indicated values in FIG. 2 (B) of the, L 2 instead be approximated by Equation 6, 7, the scattering coefficient using Equation 5 as S w' May be determined. In this case, temporarily determines the values of the scattering coefficient S B of the base glaze in step S 2 3 6, to determine the value of the scattering coefficient S B by repeatedly executing step S 2 3~S 26 do it. In this way, the unknown becomes the scattering coefficient S B
1つだけなので、 ステップ S 2 3において散乱係数 SB を修正するための逐次近 似法の計算の収束が早くなるという利点がある。 Since there is only one, there is an advantage that the convergence of the calculation of the successive approximation method for correcting the scattering coefficient S B in step S 23 becomes faster.
ただし、 ステップ S 2 3では、 逐次近似法によって係数を求める必要はなく、 種々の係数の値を仮定して、 ステップ S 26におけるシミュレーション結果が最 適となる係数を選択するようにしてもよレ、。 例えば、 数式 5をそのまま用いる場 合には、 Ss の値を約 0. 00 1〜約0. 0 1 0の範囲で数点設定しておき、 ス テツプ S 26において実調合率に最も近い調合率を予測できる SB の値を決定す るようにしてもよい。 However, in step S23, it is not necessary to find the coefficients by the successive approximation method, and it is possible to assume various coefficient values and select the coefficient that optimizes the simulation result in step S26. ,. For example, when using Equation 5 as it is, set several values of S s in the range of about 0.001 to about 0.010, and in Step S26, the value closest to the actual mixing rate is set. the compounding ratio may be so that to determine the value of S B predictable.
散乱係数 S を表現するための関数は数式 5〜 7に限られるわけではなく、 一般に、 散乱係数 Sw' を調合率 Cw に依存した関数 f (Cw ) で表現するよう にすればよい。 The function for expressing the scattering coefficient S is not limited to Equations 5 to 7.In general, the scattering coefficient S w ′ may be expressed by a function f (C w ) that depends on the mixing ratio C w .
なお、 上記第 1実施例では、 着色剤の調合率を外添加調合率 (被着色物の調合 率 SB を 1 00 %とした時の着色剤の調合率) で定義したが、 調合率を他の規定 の仕方で定義することも可能である。 従って、 数式 5は、 定数 3l を含む次の数 式 23によって表現することができる。 In the above-described first embodiment, defining the formulation ratio of the colorant in the outer additive formulation ratio (compounding ratio of the colorant when the 1 100% The formulation ratio S B of the coloring material), the preparation rate It can be defined in other prescribed ways. Therefore, Equation 5 can be expressed by the following Equation 23 including the constant 3l .
Sw' = ^i¾ ...(23) Sw '= ^ i¾ ... (23)
し、 ν + ί  Then ν + ί
(2) ΧΥΖ表色系や L*a*b* 表色系の値を一致させるコンピュータカラー マッチングは、 一般にメタメリックマッチ法と呼ばれている。 一方、 混合物の分 光反射率 R ( λ ) の曲線を一致させるァイソメリックマッチ法と呼ばれる方法も ある。 この発明は、 メタメリックマッチ法のみでなく、 ァイソメリックマッチ法 によるコンピュータカラーマッチングにも適用可能である。 (2) コ ン ピ ュ ー タ Computer color matching that matches the values of the color system and the L * a * b * color system is generally called the metameric matching method. On the other hand, There is also a method called an isomeric matching method for matching the curves of the light reflectance R (λ). The present invention is applicable not only to the metameric matching method but also to computer color matching by the isometric matching method.
( 3 ) 上記第〗実施例では陶磁器の釉を对象とするコンピュータカラ一マツチン グについて説明したが、 本発明はこれに限らず、 他の種類の混合物を対象とする コンピュータカラーマッチングにも適用することが可能である。 但し、 陶磁器の 素地は、 完全な白色や完全な黒色のものを作成することは困難なので、 絶対法に よって被着色物 (ベース釉) 単独の物性値を決定することは難しい。 従って、 釉 に関するコンピュータカラ一マッチングに本発明を適用すれば、 予測精度を向上 させる上で特に効果が大きい。  (3) In the above-described first embodiment, computer color matching using ceramic glaze has been described. However, the present invention is not limited to this, and is also applicable to computer color matching for other types of mixtures. It is possible. However, since it is difficult to produce a completely white or completely black ceramic body, it is difficult to determine the physical properties of the object to be colored (base glaze) by the absolute method. Therefore, applying the present invention to computer color matching for glaze is particularly effective in improving prediction accuracy.
以下では, 第 1実施例で求められた散乱係数と吸収係数とを用いたコンビユー タカラーマッチングの種々の適用例としての各種の実施例を説明する。 但し, 以 下の各実施例は, 第 1実施例で決定された散乱係数と吸収係数とを用いる場合に 限られるわけではなく, 他の方法で決定された散乱係数と吸収係数とを用いるこ とも可能であることに注意すべきである。  In the following, various embodiments will be described as various application examples of the computer color matching using the scattering coefficient and the absorption coefficient obtained in the first embodiment. However, each of the following embodiments is not limited to the case where the scattering coefficient and the absorption coefficient determined in the first embodiment are used, and the scattering coefficient and the absorption coefficient determined by other methods are used. It should be noted that both are possible.
Β . 第 2実施例: Β. Second embodiment:
一般に、 コンピュータカラ一マッチングによる予測には誤差があるので、 その 予測誤差を小さくする工夫が である。 従来は、 予測誤差を減少させるために、 各着色剤の吸収係数 Ki と反射係数 S i とを正確な値に近づけるように補正して いた。  In general, there is an error in the prediction by computer color matching, so a method to reduce the prediction error is proposed. Conventionally, in order to reduce the prediction error, the absorption coefficient Ki and the reflection coefficient S i of each colorant have been corrected so as to be close to accurate values.
し力 し、 各成分の吸収係数 と反射係数 S i とを正確な値に近づけるた めには、 被着色物に各着色剤を単独で舍む数多くの混合物サンプルを作成して、 その分光反射率を測定しなければならず、 膨大な作業を要していた。 また、 天然 の顔料を着色剤として用いる場合には、 その吸収係数 κ± や散乱係数 S i が必ず しも一定の値にはならないので、 吸収係数 Ki と散乱係数 S i を正確な値に近づ けることが困難であった。 In order to make the absorption coefficient and the reflection coefficient S i of each component close to the exact values, a large number of mixture samples were prepared, in which each colorant was added to the object to be colored, and the spectral reflection The rate had to be measured and required a lot of work. When a natural pigment is used as a coloring agent, the absorption coefficient κ ± and the scattering coefficient S i do not always become constant, so that the absorption coefficient Ki and the scattering coefficient S i are close to accurate values. It was difficult to find out.
以下に示す第 2実施例は、 各成分の吸収係数 Ki と散乱係数 S i とを補正する ことなく、 予測誤差を減少させることを目的としたものである。 The second embodiment described below corrects the absorption coefficient Ki and the scattering coefficient S i of each component. The aim is to reduce the prediction error without any change.
図】 1は、 第 2実施例における処理の全体手順を示すフローチャートである。 なお、 この第 2実施例で対象とする混合物は、 陶磁器の素地の表面を覆うための 釉 (ゆう) である。 すなわち、 顔料を入れないベース釉 (基礎釉) が被着色物で あり、 このべ一ス釉に顔料を添加した釉がコンピュータカラ一マッチングの対象 となる混合物である。  FIG. 1 is a flowchart illustrating an overall procedure of a process according to a second embodiment. The mixture targeted in the second embodiment is a glaze for covering the surface of the ceramic body. In other words, the base glaze (base glaze) without pigment is the object to be colored, and the glaze obtained by adding pigment to this base glaze is the mixture to be subjected to computer color matching.
ステップ S 3 1では、 コンピュータカラ一マッチングの予測結果 (三刺激値) を補正するためのニューラルネットワークの学習を行なう。 ステップ S 3 2〜S 3 5では、 学習済みのニューラルネットワークを用いてコンピュータカラーマツ チングの目標値を補正して、 正確な予測結果を求めている。 以下ではまず、 ステツ プ S 3 2〜S 3 5の内容を説明する前に、 ニューラルネットワークの構成とステツ プ S 3 1の詳細手順について説明する。  In step S31, learning of a neural network for correcting the prediction result (tristimulus value) of computer color matching is performed. In steps S32 to S35, the target value of computer color matching is corrected using the trained neural network to obtain an accurate prediction result. Hereinafter, before describing the contents of steps S32 to S35, the configuration of the neural network and the detailed procedure of step S31 will be described first.
図 1 2は、 ニューラルネットワークの構成を示す説明図である。 この二ユーラ ルネットワークは、 入力層 1 0と中閒肩 2 0と出力層 3 0とで構成される三層の 階層構造を有している。 人力層 I 0は、 3つのニューロン N 11〜N 13で構成さ れており、 中間層 2 0は 5つのニューロン N 21〜N 25で、 出力肩 3 0は 3つの ニューロン N31〜N33でそれぞれ構成されている。  FIG. 12 is an explanatory diagram showing the configuration of the neural network. This dual-purpose network has a three-layer hierarchical structure composed of an input layer 10, a middle shoulder 20 and an output layer 30. The human stratum I0 is composed of three neurons N11 to N13, the middle layer 20 is composed of five neurons N21 to N25, and the output shoulder 30 is composed of three neurons N31 to N33. Have been.
入力層 1 0の 3つのニューロン N 11〜N 13には、 二:刺激値 X, Υ , Zがそれ ぞれ入力される。 入力層 1 0のニューロン N ijから中間層 2 0のニューロン Nk に伝達される信号は、 それぞれの入力信号に重み W , k を乗じたものである。 ここで、 iは注目している階屑を示す番号、 jは注目している階層内でのニュー ロンの順番を示す番号、 kは次の階層のニューロンの順番を示す番号である。 例 えば、 入力餍 1 0の第 1のニューロン N ilから中聞層 2 0の第 1 のニューロン N 21に伝達される信号は Wu. であり、 入力屑 1 0の第 1のニューロン N ilか ら中間層 2 0の第 2のニューロン N 22に伝達される信号は Wu, 2Xである。 Two: stimulus values X, ,, and Z are input to the three neurons N11 to N13 of the input layer 10 respectively. The signal transmitted from the neuron Nij of the input layer 10 to the neuron Nk of the hidden layer 20 is obtained by multiplying each input signal by weights W and k . Here, i is a number indicating the debris of interest, j is a number indicating the order of neurons in the layer of interest, and k is a number indicating the order of neurons in the next layer. For example, the signal transmitted from the first neuron N il of the input 餍 10 to the first neuron N 21 of the middle layer 20 is W u ., And the first neuron N il of the input dust 10 is signal transmitted to the second neuron N 22 of pressurized et intermediate layer 2 0 is W u, 2 X.
中聞層 2 0の各ニューロン N ijの入力 と出力 Q Ajとの関係は、 次の数式 2 4に示す情報伝達関数 f ( U ij) で与えられる。
Figure imgf000031_0001
The relationship between the input and output Q Aj of each neuron N ij in the middle hearing layer 20 is given by an information transfer function f ( U ij ) shown in the following Expression 24.
Figure imgf000031_0001
ここで、 は ( i 一〗) 番目の階層 (すなわち入力層 1 0) の〗番目の ニューロン N (i- l)jの出力であり、 闵 1 2の例では QU=X, 012=Y, 013 = Ζである。 また、
Figure imgf000031_0002
k は ( i— 1) 番目の階層の j番目のニューロン N (i-1 ) jから、 注目しているニューロン Nijに伝達される信号に掛かる重みであ る。 tはしきい値であり、 一定の他が割当てられる。 なお、 数式 24の情報伝達 関数 f (u , はシグモイド関数と呼ばれている。
Here, is the output of the〗 th neuron N (i-l) j in the (i〗 1) th hierarchy (that is, the input layer 10). In the example of 闵12 , Q U = X, 0 12 = Y, 0 13 = Ζ. Also,
Figure imgf000031_0002
k is the weight applied to the signal transmitted from the j-th neuron N (i-1) j in the (i-1) th hierarchy to the neuron Nij of interest. t is a threshold, some other being assigned. Note that the information transfer function f (u, in Equation 24 is called a sigmoid function.
例えば、 数式 24を中閒層 20の第〗のニューロン N21の入出力関係に適用 すると、 次の数式 25が得られる。
Figure imgf000031_0003
u2i = Wn,iX + W,2jY + W131Z -..(25b) 出力層 30の各ニューロンの入出力関係も上記数式 24で与えられる。 図 1 2 に示す第 2実施例では、 出力層 30の 3つのニューロン N31〜N33の出力
For example, when Equation 24 is applied to the input / output relationship of the second neuron N21 of the middle layer 20, the following Equation 25 is obtained.
Figure imgf000031_0003
u 2 i = W n , iX + W, 2jY + W 131 Z-.. (25b) The input / output relationship of each neuron of the output layer 30 is also given by the above equation (24). In the second embodiment shown in FIG. 12, the output of the three neurons N31 to N33 in the output layer 30 is
〜Q13を、 コンピュータカラ一マッチング (CCM) による三刺激値の予測誤差 ΔΧ, ΔΥ, ΔΖとしている。 The to Q 13, predicts errors ΔΧ tristimulus values by computer color one matching (CCM), ΔΥ, and Derutazeta.
ニューラルネットワークの学習は、 入力屑 1 0への入力 (X, Y, Z) と出力 層 30からの出力 (ΔΧ, ΔΥ, ΔΖ) との間の関係を数多く与えて、 正しい入 出力関係を与えるような重み W , k の値を決定する作業である。 Neural network learning gives a number of relationships between the inputs (X, Y, Z) to the input debris 10 and the outputs (ΔΧ, ΔΥ, ΔΖ) from the output layer 30 to give the correct input / output relationships. This is the task of determining the values of the weights W and k .
図 1 3は、 図 1 1のステップ S 3 1の詳細手順を示すフローチャートである。 ステップ S 4 1では、 コンピュータカラ一マッチングにおいて予測の対象とする 混合物の色の範囲をカバーするような複数の三刺激値 (Xit. Yit, Zit) 定めて、 これらの複数の三刺激値を有するような複数のサンプルの調合割合をコ ンピュータカラーマッチングで決定する。 図 1 4は、 コンピュータカラーマッチ ングでの色の予測対象範囲 P Aと、 この予測対象範囲 P Aをカバーする複数の三 刺激値の分布を示す概念図である。 この第 2実施例では C I E— XYZ表色系で 色を表現するものとしており、 色の予測対象範囲 P Aは X Y Z座標系の 3次元的 な範囲として^えられる。 なお、 予測対象範囲 P Aは、 予測対象とする混合物が 取り得る色の範囲を示すものであり、 任意に設定し得る範囲である。 FIG. 13 is a flowchart showing the detailed procedure of step S31 in FIG. In step S41, a plurality of tristimulus values (Xit. Yit, Zit) that cover the color range of the mixture to be predicted in computer color matching Then, the blending ratio of a plurality of samples having the plurality of tristimulus values is determined by computer color matching. FIG. 14 is a conceptual diagram illustrating a color prediction target range PA in computer color matching and a distribution of a plurality of tristimulus values covering the prediction target range PA. In the second embodiment, colors are expressed in the CIE-XYZ color system, and the color prediction target range PA can be obtained as a three-dimensional range in the XYZ coordinate system. Note that the prediction target range PA indicates the range of colors that the mixture to be predicted can take, and is a range that can be set arbitrarily.
この第 2実施例においては、 予測対象範囲 P Aをカバーするために、 図〗 4に 白丸で示す 7組の三刺激値を決定した。 ステップ S 4 1では、 さらに、 これらの 7組の三刺激値 (Xit, Y it, Z it) を有するような 7種類のサンプル M 1〜 M 7の調合割合を、 コンピュータカラーマッチングによって予測した。  In the second example, seven sets of tristimulus values indicated by white circles in FIG. 4 were determined in order to cover the prediction target range PA. In step S41, the compounding ratio of the seven types of samples M1 to M7 having these seven sets of tristimulus values (Xit, Yit, Zit) was further predicted by computer color matching.
ここで、 コンピュータカラーマッチングによる色予測と調合割合の予測につい て簡単に説明する。 第 1実施例で説明した数式 2を変形すると、 混合物の分光反 射率 R (ス) は、 次の数式 2 6で与えられる。  Here, the color prediction by computer color matching and the prediction of the mixing ratio will be briefly described. By transforming Equation 2 described in the first embodiment, the spectral reflectance R (s) of the mixture is given by Equation 26 below.
R= 1 + fK] ffi IK} ...(26) R = 1 + fK] ffi IK} ... (26)
M 混合物の吸収係数と散乱係数の比 (KZS) κは、 各成分の吸収係数 Ki (1) と散乱係数 S± ) と調合率 Ci から, 第 1実施例で説明した数式 1に従って 算出できるので、 混合物の分光反射率 R (λ ) を上記数式 2 6から求めることが できる。 この分光反射率 R (λ ) は理想状態 (被着色物の厚みが無限大の場合) の分光反射率なので、 分光光度計で測定できる分光反射率 R' (; -) を、 次の数 式 2 7 (サンダーソンの式) に従って求める。 The ratio (KZS) κ between the absorption coefficient and the scattering coefficient of the M mixture can be calculated from the absorption coefficient Ki (1) and the scattering coefficient S ± ) of each component and the mixing ratio Ci according to Equation 1 described in the first embodiment. The spectral reflectance R (λ) of the mixture can be obtained from the above equation 26. Since this spectral reflectance R (λ) is the spectral reflectance in an ideal state (when the thickness of the object to be colored is infinite), the spectral reflectance R '(;-) that can be measured by a spectrophotometer is expressed by the following equation. 2 7 (Sanderson's equation)
R' = k1-Kl -ki)(l -k2)l ..- (27) ここで、 係数 , k 2 は、 被着色物 (ベース釉) の光学的性質に依存する 値である。 係数 , k 2 としては、 被着色物の屈折率 nから次の数式 2 8に 従って決定することができる。
Figure imgf000033_0001
R '= k 1 -Kl -ki) (l -k 2 ) l ..- (27) where the coefficient, k 2 is a value that depends on the optical properties of the object to be colored (base glaze). The coefficients, k 2 can be determined from the refractive index n of the object to be colored according to the following equation 28.
Figure imgf000033_0001
k2=0.68n-0.56 . - . (28b) k 2 = 0.68n-0.56 .-. (28b)
なお、 第 2実施例において用いたベース釉は、 屈折率 nが約 1. 4である。 数式 27によって分光反射率 R' ) が求まると、 混^の三刺激値 X, Y, Zが次の数式 29によって求められる。  The base glaze used in the second example had a refractive index n of about 1.4. When the spectral reflectance R ') is determined by Equation 27, the tristimulus values X, Y, and Z of the mixture are determined by Equation 29 below.
X = k]S( )R'( )x( )dX ...(29a) X = k] S () R '() x () dX ... (29a)
Y = kj S( )R'( )y( )dX - - . (29b) Y = kj S () R '() y () dX--. (29b)
Z = k)S(X)R )z^)dX ... (29c) k =,—— ... (29d) Z = k) S (X) R) z ^) dX ... (29c) k =, ---- ... (29d)
ここで、 S (λ ) は標準光の分光分布、 X (/) , y (λ ) , ζ (λ ) (数式 中ではバー付きである) は等色関数である。 Here, S (λ) is the spectral distribution of the standard light, and X (/), y (λ), and ζ (λ) (with a bar in the formula) are color matching functions.
コンピュータカラ一マッチングによって混合物の色を予測する場合には、 上述 のように、 混合物の各成分の吸収係数と散乱係数に基づき、 数式 26〜29 に従って、 その混合物の三刺激値 X, Υ, Ζを算出する。 三刺激値は混合物の色 を表わすので、 任意の混合物の色を予測することができることになる。  When predicting the color of a mixture by computer color matching, as described above, based on the absorption coefficient and scattering coefficient of each component of the mixture, the tristimulus values X, Υ, Ζ Is calculated. Since the tristimulus values represent the color of the mixture, the color of any mixture can be predicted.
また、 所望の色を有する混合物の調合割合を予測する場合には、 混合物の調合 割合を仮定して、 上述した手順でその三刺激値を算出し、 ニュートン一ラプソン 法などの逐次近似法によって所望の色に所定の誤差内で一致するような調合割 1 を求める。  Further, when predicting the blending ratio of a mixture having a desired color, the tristimulus value is calculated by the above-described procedure assuming the blending ratio of the mixture, and the desired value is calculated by a successive approximation method such as the Newton-Raphson method. Formula 1 is determined so that the color matches the color within a specified error.
図 1 3のステップ S 42では、 ステップ S 4 〗で予測された調合割合 C iを有 する複数のサンプルを作成する。 この第 2実施例では、 図 1 4の 7組の三刺激値 (Xit, Yit, Zit) に対応する 7つのサンプル M 1〜M 7を作成した。 ステツ プ S 43では、 各サンプル M iの分光反射率を分光光度計で測定し、 上記数式 2 9に従ってその三刺激値 (Xim Yim, Zin を求める。 ステップ S 44では、 各サンプル M iに関して、 ステップ S 43で得られた実 測値 (Xim, Yim, Zim) とステップ S 4 1で決定した目標値 (Xit, Yit, Zit) との差をとることによって、 予測誤差 AMi (Xim-Xit, Yim-Yit- , Zim— Zit) を求める。 図 1 5は、 第 2実施例において得られた 7つのサン プル M 1〜M 7の予測誤差を示す概念図である。 また、 図 1 6は、 各サンプルの 三刺激値の目標値 (ステップ S 4 1で決定された値) と予測誤差 ΔΜΪ (ΔΧ, ΔΥ, ΔΖ) を示す説明「¾1である。 In step S42 of FIG. 13, a plurality of samples having the compounding ratio C i predicted in step S4〗 are created. In the second embodiment, seven samples M1 to M7 corresponding to the seven sets of tristimulus values (Xit, Yit, Zit) in FIG. 14 were created. In step S43, the spectral reflectance of each sample Mi is measured with a spectrophotometer, and its tristimulus values (Xim Yim, Zin) are obtained according to the above equation (29). In step S44, for each sample Mi, the difference between the measured value (Xim, Yim, Zim) obtained in step S43 and the target value (Xit, Yit, Zit) determined in step S41 is calculated. By doing so, the prediction error AMi (Xim-Xit, Yim-Yit-, Zim—Zit) is obtained. FIG. 15 is a conceptual diagram illustrating prediction errors of seven samples M1 to M7 obtained in the second example. FIG. 16 is a description “説明 1” showing the target value of the tristimulus value (the value determined in step S41) of each sample and the prediction error ΔΜΪ (ΔΧ, ΔΥ, ΔΖ).
なお、 ステップ S 4 1で得られた調合割合 C iからコンピュータカラ一マッチ ングによって各サンプルの三刺激俩の予測値 (Xic, Yic, Zic) を求め、 実 測値 (Xim, Yim, Zim) と予測愤 (Xic, Yic, Zic) との差 (Xim—X ic, Y im-Y ic, Zim— Zic) を予測誤差 AMiと定義してもよい。 コン ピュータカラーマッチングで調合割合 C iを決定する際には、 三刺激値の冃標値 The predicted values (Xic, Yic, Zic) of the tristimulus の of each sample were obtained from the mixture ratio C i obtained in step S41 by computer color matching, and the measured values (Xim, Yim, Zim) were obtained. The difference (Xim—Xic, Yim-Yic, Zim—Zic) between the prediction error and the prediction 愤 (Xic, Yic, Zic) may be defined as the prediction error AMi. When determining the mixing ratio C i by computer color matching, the target value of tristimulus values
(Xit, Yit, Zit) と予測値 (Xic, Yic, Zic) との差が所定の許容誤 差以下になるように調合割合 C iを決定するので、 目標値 (Xit, Yit, Zit- ) と予測値 (Xic, Yic, Zic) は実質的にほぼ等しい値を有している。 従つ て、 実測値 (Xim, Yim, Zim) と予測愤 (Xic, Yic, Zic) との差 (X im-Xic, Yim— Yic, Zim— Zic) を予測誤差 ΔΜίと定義しても、 実測値Since the mixing ratio C i is determined so that the difference between (Xit, Yit, Zit) and the predicted value (Xic, Yic, Zic) is equal to or less than a predetermined allowable error, the target value (Xit, Yit, Zit-) And the predicted values (Xic, Yic, Zic) have substantially the same value. Therefore, even if the difference (X im-Xic, Yim—Yic, Zim—Zic) between the measured value (Xim, Yim, Zim) and the prediction 愤 (Xic, Yic, Zic) is defined as the prediction error ΔΜί, Measured value
(Xim, Yim, Zim) と目標値 (Xit, Yit, Zit) との差 (Xim-Xit, Yim- Yit, Zim- Zit) を予測誤差 AMiと定義しても実質的にはほぼ同じで ある。 Even if the difference (Xim-Xit, Yim-Yit, Zim-Zit) between (Xim, Yim, Zim) and the target value (Xit, Yit, Zit) is defined as the prediction error AMi, it is virtually the same. .
図 1 3のステップ S 45では、 図 1 6に示す各サンプル M iの三刺激値の目標 値 (Xit, Yit, Zit) と予測誤差 ΔΜΐとを用いてニューラルネットワークの 学習を行ない、 上記数式 24における重み Wij, kを決定する。 ニューラルネッ トワークの学習方法としては、 例えば逆誤差伝搬学習方式を用いる。  In step S45 in FIG. 13, learning of the neural network is performed using the target values (Xit, Yit, Zit) and the prediction error Δ サ ン プ ル of the tristimulus values of each sample Mi shown in FIG. Determine the weights Wij, k at. As a learning method of the neural network, for example, a back error propagation learning method is used.
図 1 7は、 第 2実施例におけるニューラルネットワークの学習の実証結果を示 す説明図である。 ここでは、 図 1 4, 5に示す予測対象範囲 P Aに含まれるもう 1組の三刺激値をコンピュータカラ一マッチングの目標値として設定し、 この目 標値を有する第 8のサンプル M 8を作成した。 そして、 このサンプル M 8の三刺 激俯を実測した。 図 I 7において 「CCM目標値」 とあるのは第 8のサンプル M 8のコンピュータカラーマッチングに用いた目標値を意味している。 また、 「誤 差 (真植) 」 とあるのは三刺激値の目標値と実測値との差である。 「ニューロ予 測誤差」 は、 CCM目標値を学習済みのニューラルネットワーク (図 1 2) に入 力した場合に得られる予測誤差である。 図〗 7の結果から、 学習済みの二ユーラ ルネットワークは、 三刺激値の誤差を精度良く予測できることが解る。 FIG. 17 is an explanatory diagram showing the results of the verification of the neural network learning in the second embodiment. Here, another set of tristimulus values included in the prediction target range PA shown in Figs. 14 and 5 is set as the target value for computer color matching. An eighth sample M8 having a standard value was prepared. Then, the sagittal depression of this sample M8 was actually measured. In FIG. I7, “CCM target value” means the target value used for the computer color matching of the eighth sample M8. In addition, "error (true planting)" is the difference between the target value and the measured value of the tristimulus value. “Neuro prediction error” is the prediction error obtained when the CCM target value is input to a trained neural network (Fig. 12). From the results in Fig. 7, it can be seen that the trained bi-ural network can accurately predict the error of tristimulus values.
こうしてニューラルネットワークの学習が終了すると、 図 1 1のステップ S 3 2〜S 35を実行して、 調合率未知の色見本の調合率を予測する。 ステップ S 3 2では、 調合率未知の色見本の分光反射率を測定してその三刺激値 (X s, Y s, Z s ) を求める。 ステップ S 3 3では、 図 1 2に示すニューラルネットワークに 色見本の三刺激値 (X s, Y s, Z s) を入力して、 予測誤差△ (AX s, AY s, ΔΖ s) を求める。 テツプ S 34では、 三刺激値 (X s, Y s, Z s ) を予 測誤差厶で補正して、 コンピュータカラーマッチングの調合割合予測に用いる三 刺激値の目標储 (X s— ΔΧ s, Y s -ΔΥ s , Zs—AZs) を求める。 ステツ プ S 35では、 こうして補正された目標値を用いてコンピュータカラ一マツチン グを実行し、 色見本の調合率を予測する。  When learning of the neural network is completed in this way, steps S32 to S35 in FIG. 11 are executed to predict the mixing ratio of a color sample whose mixing ratio is unknown. In step S32, the spectral reflectance of the color sample whose mixing ratio is unknown is measured, and its tristimulus value (Xs, Ys, Zs) is obtained. In step S33, the tristimulus values (Xs, Ys, Zs) of the color samples are input to the neural network shown in Fig. 12, and the prediction error △ (AXs, AYs, ΔΖs) is obtained. . In step S34, the tristimulus values (Xs, Ys, Zs) are corrected by the prediction error, and the target of the tristimulus values 储 (Xs—ΔΧs, Y s -ΔΥ s, Zs—AZs). In step S35, computer color matching is performed using the corrected target value to predict the mixing ratio of the color sample.
図 1 8は、 第 2実施例におけるコンピュータカラーマッチングの予測精度を検 証するために行なった実験結果を示す説明図である。 ここでは、 予測精度を実証 することを目的としたので、 調合率既知の色見本について三刺激値を実測し、 そ の実測値 (X s, Y s, Z s) を実現する調合率をコンピュータカラーマツチン グで予測した。 図 1 7の結果から、 調合率の真値とコンピュータカラーマツチン グによる予測値とは極めて良く一致していることが解る。  FIG. 18 is an explanatory diagram showing the results of an experiment performed to verify the prediction accuracy of computer color matching in the second embodiment. Here, we aimed to verify the prediction accuracy, so we measured the tristimulus values for a color sample with a known blending ratio, and calculated the blending ratio that realized the measured values (Xs, Ys, Zs) using a computer. Predicted by color matching. From the results in Fig. 17, it can be seen that the true value of the blending ratio and the predicted value by computer color matching are in very good agreement.
なお、 図 1 2に示す学習済みのニューラルネットワークを利用すれば、 調合率 既知の混合物の三刺激値も精度良く予測することが可能である。 すなわち、 予測 対象の混合物の調合率 C iから上記数式】, 2 6〜 29を用いて三刺激値を求め、 この三 撒値を図 1 2のニューラルネッ卜ワークに入力して予測誤差△を求める。 そして、 数式 29で得られた三刺激値を予測誤差厶で補正すれば、 実際の値に極 めて近い :刺激値が得られる。 If the learned neural network shown in FIG. 12 is used, it is possible to accurately predict the tristimulus values of the mixture with a known mixing ratio. That is, tristimulus values are obtained from the mixture ratio C i of the mixture to be predicted using the above formulas, 26 to 29, and the tristimulus values are input to the neural network shown in FIG. Ask. Then, if the tristimulus value obtained by Equation 29 is corrected by the prediction error, it will be extremely close to the actual value. Very close: Stimulus value is obtained.
プラスチック等と異なり, 陶磁器においては, 焼成条件, 熔融中の原料の化学 反応等により, 混色理論が成立しない場合がある。 この場合, ニューラルネット ワークに混合物の三刺激値と C C Mによる予測誤差の関係を学習させる本実施例 の方法は有効である。  Unlike plastics, etc., in ceramics, the theory of color mixing may not be established due to firing conditions, chemical reaction of raw materials during melting, and the like. In this case, the method of this embodiment in which the neural network learns the relationship between the tristimulus value of the mixture and the prediction error due to CCM is effective.
なお、 上記第 2実施例に関しては、 例えば次のような変形も可能である。  Note that, for the second embodiment, for example, the following modifications are also possible.
( 1 ) 上記第 2実施例ではコンピュータカラーマッチングにおける三刺激値の目 標値をニューラルネットワークの入力としていたが、 ニューラルネットワークの 人力としては、 各サンプル M iの三刺激値の実測値 (Xim, Yim, Zim) を用 いても良く、 また、 コンピュータカラーマッチングによる予測値 (Xic, Yic- , Zic) を用いても良い。 すなわち、 ニューラルネットワークには、 複数のサ ンプルの三刺激値 (XYZ表色系の座標値) とその予測誤差との関係を学習させ るようにすればよい。  (1) In the second embodiment, the target values of the tristimulus values in the computer color matching were input to the neural network. However, as the human power of the neural network, the measured tristimulus values (Xim, Yim, Zim) may be used, or predicted values (Xic, Yic-, Zic) by computer color matching may be used. In other words, the neural network may be made to learn the relationship between the tristimulus values (coordinate values in the XYZ color system) of a plurality of samples and their prediction errors.
(2) t記第 2実施例では、 ニューラルネットワークを用いて三刺激値を補正し ていたが、 回帰分析やニューロフアジィ技術等の他の誤差補正法を用いて補正す ることも可能である。  (2) In the second embodiment, the tristimulus values are corrected using a neural network, but may be corrected using other error correction methods such as regression analysis or neuro-fuzzy technology. is there.
(3) 上記第 2実施例では XYZ表色系を用いていたが、 本発明は、 L*a*b* 表色系などの他の任意の表色系を用いる場合に適用できる。  (3) Although the XYZ color system is used in the second embodiment, the present invention can be applied to the case where any other color system such as the L * a * b * color system is used.
(4) XYZ表色系や L*a*b* 表色系の座標値を一致させるコンピュータ力 ラーマッチングは、 一般にメタメリックマッチ法と呼ばれている。 一方、 混合物 の分光反射率 R (λ ) の曲線を一致させるァイソメリックマッチ法と呼ばれる方 法もある。 この発明は、 メタメリックマッチ法のみでなく、 ァイソメリックマツ チ法によるコンピュータカラーマッチングにも適用可能である。  (4) Computer color matching that matches the coordinate values of the XYZ color system and the L * a * b * color system is generally called the metameric matching method. On the other hand, there is also a method called an isomeric matching method in which a curve of the spectral reflectance R (λ) of a mixture is matched. The present invention is applicable not only to the metameric matching method but also to the computer color matching by the isometric matching method.
( 5 ) 上記第 2実施例では陶磁器の釉を対象とするコンピュータカラーマツチン グについて説明したが、 本発明はこれに限らず、 他の種類の混合物を対象とする コンピュータカラーマッチングにも適用することが可能である。 C. 第 3実施例: (5) In the second embodiment, computer color matching for ceramic glaze was described, but the present invention is not limited to this, and is also applicable to computer color matching for other types of mixtures. It is possible. C. Third Embodiment:
衛生陶器などのような工業陶器には、 ユーザの様々な好みに応じて種々の色が 設定されている。 陶器の色は、 釉に混合する顔料の調合率によって決定されるの で、 釉は、 コンピュータカラーマッチングにおける混合物に相当する。 陶器の色 を設定する際には、 まず、 設計者が紙などを彩色したり、 予め焼成されたサンプ ルから所望の色を有するサンプルを選択したりすることによって色を決定する。 そして、 その色の分光反射率を分光光度計で測定し、 分光反射率の実測値からコ ンピュータマツチングを用いて顔料や染料等の着色剤の調合割合を予測する。 ところで、 同じ調合割合の釉薬を用いても、 製造される陶器の実際の色は同一 にはならず、 色にばらつきがあるのが普通である。 図〗 9は、 設計された色と実 際に製造される陶器の色のばらつきを示する X— y色度図である。 図 1 9におい て、 二重丸で示される色が設計された所望の色 (標準色) L 0であり、 小さな白 丸が実際に製造される陶器の色の分布である。 標準色 L 0から大きく異なる色を 有する陶器は不良品として認識する必要がある。 そこで、 良品の色の濃淡の限度 を示すものとして、 図 1 9に黒丸で示す色を有する濃淡限度色 L 1, L 2が設定 される。 標準色 L 0の色度座標値は設計者によって決定されているので、 この標 準色 L 0の色度座標値に製造誤差等を考慮して、 2つの'濃淡限度色 L 1, L 2の 色度座標値を決定する。  Industrial porcelain, such as sanitary ware, has various colors set according to various user preferences. Since the color of the pottery is determined by the blending ratio of the pigment mixed with the glaze, the glaze corresponds to the mixture in computer color matching. When setting the color of the pottery, the designer first determines the color by coloring the paper or the like or selecting a sample having a desired color from pre-baked samples. Then, the spectral reflectance of the color is measured with a spectrophotometer, and the mixing ratio of a coloring agent such as a pigment or a dye is predicted from the measured value of the spectral reflectance using computer matching. By the way, even if glazes with the same blending ratio are used, the actual colors of the pottery produced will not be the same, and the colors will usually vary. Figure 9 is an X-Y chromaticity diagram showing the variation between the designed colors and the colors of the pottery actually manufactured. In Fig. 19, the color indicated by the double circle is the designed color (standard color) L0, and the small white circle is the color distribution of the ceramics actually manufactured. Pottery with a color that is significantly different from the standard color L0 must be recognized as defective. Therefore, the shade limit colors L1 and L2, which have the colors indicated by the black circles in Fig. 19, are set to indicate the shade limits of non-defective products. Since the chromaticity coordinate values of the standard color L0 are determined by the designer, the chromaticity coordinate values of the standard color L0 are taken into consideration by considering the manufacturing error and the like, and the two 'shade limit colors L1, L2' Determine the chromaticity coordinate value of.
実際の陶器の製造においては、 標準色 L 0を有する陶器を標準色見本として作 成し、 また、 濃淡限度色 L I , L 2を有する 2つの陶器を濃淡限度見本として作 成する。 そして、 検査工程において、 標準色見本および渙淡限度見本の色と製造 された陶器の色とを比較して、 渙淡限度見本の色の範囲にある陶器のみを良品と する。  In the actual production of pottery, a pottery having the standard color L0 is created as a standard color sample, and two potteries having shading limit colors L I and L2 are created as shading limit samples. Then, in the inspection process, the color of the standard color sample and the color limit sample is compared with the color of the manufactured ceramic, and only the ceramics in the color range of the color limit sample are regarded as good products.
ところで、 コンピュータカラ一マッチングによる予測にはかなりの誤差が伴う ので、 予測された調合割合から標準色見本ゃ澹淡限度見本を作成するにはかなり の手間を要するのが普通である。 例えば、 標準色見本を作成する際には、 コン ピュータカラーマッチングで予測された調合割合で調合した釉薬を用いて多数の サンプルを焼成し、 その中で所望の標準色を有するサンプルがあればそれを標準 色見本として採用する。 しかし、 標準色を有するサンプルが無ければ、 勘と経験 に従って調合割合を変更して再度多数のサンプルを焼成し、 標準色を有するサン プルが得られるまでこの作業を繰り返す。 このような作業は、 濃淡限度色 L 1, L 2を有する濃淡限度見本についても行なわれる。 By the way, since the prediction by the computer color matching involves a considerable error, it usually takes a considerable amount of time to create a standard color sample and a shading limit sample from the predicted mixing ratio. For example, when preparing a standard color sample, a large number of samples are fired using glaze prepared at the mixing ratio predicted by computer color matching, and if there is a sample having the desired standard color, The standard Adopt as a color sample. However, if there is no sample having the standard color, the mixing ratio is changed according to intuition and experience, and a large number of samples are baked again, and this operation is repeated until a sample having the standard color is obtained. Such an operation is also performed on the gray scale sample having the gray scale colors L1 and L2.
このように、 従来のコンピュータカラーマッチングでは、 所望の色を有する混 合物を作成するための調合割合を精度良く予測するのが困難であった。  As described above, in the conventional computer color matching, it has been difficult to accurately predict the mixing ratio for producing a mixture having a desired color.
以下に示す第 3実施例は、 所望の色を有する混合物の調合割合を精度良く予測 することを目的としたものである。  The third embodiment described below aims at accurately estimating the blending ratio of a mixture having a desired color.
図 2 0は、 第 3実施例における処理の全体手順を示すフローチャートである。 なお、 この第 3実施例で対象とする混合物は、 陶磁器の素地の表面を覆うための 釉 (ゆう) である。 すなわち、 顔料を入れないベース釉 (基礎釉) が被着色物で あり、 このベース釉に顔料を添加した釉がコンピュータカラーマッチングの対象 となる混合物である。  FIG. 20 is a flowchart showing the overall procedure of the process in the third embodiment. The mixture targeted in the third embodiment is a glaze for covering the surface of a ceramic body. In other words, the base glaze without pigment (base glaze) is the object to be colored, and the glaze obtained by adding pigment to this base glaze is the mixture to be subjected to computer color matching.
ステップ S 5 1では、 まず標準色見本を作成する。 従って、 標準色見本の着色 剤の混合割合は既知である。 ステップ S 5 2では、 作成された標準色見本の分光 反射率 R' を分光光度計で測定し、 この分光反射率 R' から次の数式 30に従つ て三刺激値 X。 , Υ。 , Ζ() を算出する。 In step S51, first, a standard color sample is created. Therefore, the mixing ratio of the colorant of the standard color sample is known. In step S52, the spectral reflectance R 'of the created standard color sample is measured with a spectrophotometer, and the tristimulus value X is calculated from the spectral reflectance R' according to the following equation (30). , Υ. , Ζ Calculate () .
Xo = klS(X)R*( )x( )dX - .. (30a) Xo = klS (X) R * () x () dX-.. (30a)
Y0 = kjS(X)R*(X)y(X)dX ... (30b) Zo=k(S( )R'( )z( )dX ... (30c) k = 100 . - . (30d) Y 0 = kjS (X) R * (X) y (X) dX ... (30b) Zo = k (S () R '() z () dX ... (30c) k = 100 .-. (30d)
ここで、 S (λ ) は標準-光の分光分布、 X (λ) , y ) , ζ ( Ο (数式 中ではバー付きである) は等色関数である。 Here, S (λ) is the standard-light spectral distribution, and X (λ), y) and ζ (Ο (with a bar in the formula) are color matching functions.
ステップ S 5 3では、 ステップ S 5 2で得られた標準色の三刺激値の実測値 X 。 , Υ。 , Z。 に基づいて、 製造誤差を考慮して、 設計者が図〗 9に示す 2つ の濃淡限度色 L I, L 2に対する三刺激値を設定する。 なお、 以下では説明の便 宜 、 第〗の濃淡限度色 L〗を実現するための調合割合を予測する場合について 説明する。 第 2の濃淡限度色 L 2に対しても、 同様な処理によってその調合割合 を精度良く予測することができる。 In step S53, the measured value X of the tristimulus value of the standard color obtained in step S52. , Υ. , Z. Based on the above, considering the manufacturing error, the designer Set the tristimulus values for the shade limit colors LI, L2 of. In the following, for convenience of description, a case will be described in which a blending ratio for realizing the first shade limit L 限度 is predicted. With respect to the second shade limit color L2, the blending ratio can be accurately predicted by the same processing.
ステップ S 5 4では、 ステップ S 5 1で作成した標準色見本の各着色剤の調合 率 Ci ( iは着色剤の番号を示す) と、 散乱係数 Si と、 吸収係数 Ki とから、 標準色見本の三刺激値の計算値 XM , YM , ZN を以下の手順で算出する。 まず、 第 1実施例で説明した数式 2を変形すると、 混合物の分光反射率 R (λ ) は、 次の数式 3 1で与えられる。
Figure imgf000039_0001
混合物の吸収係数と散乱係数の比 (KZS) „は、 各着色剤の吸収係数 (λ ) と散乱係数 Si ( ) と調合率 Ci から、 第 1実施例で説明した数式 1に 従って算出できる。 従って、 混合物の分光反射率 R (λ ) は、 この比 (KZS) Μ から、 数式 3 1に従って求めることができる。 この分光反射率 R (/) は理想 状態 (被着色物の厚みが無限大の場合) の分光反射率なので、 分光光度計で測定 できる分光反射率 R' (λ ) を、 次の数式 32 (サンダーソンの式) に従って求 める。
In step S54, the standard color sample is obtained from the mixing ratio Ci (i indicates the number of the colorant), the scattering coefficient Si, and the absorption coefficient Ki of each colorant of the standard color sample created in step S51. The calculated values X M , Y M , and Z N of the tristimulus values are calculated by the following procedure. First, by transforming Equation 2 described in the first embodiment, the spectral reflectance R (λ) of the mixture is given by the following Equation 31.
Figure imgf000039_0001
The ratio (KZS) 散乱 between the absorption coefficient and the scattering coefficient of the mixture can be calculated from the absorption coefficient (λ), the scattering coefficient Si (), and the blending ratio Ci of each colorant according to Equation 1 described in the first embodiment. Therefore, the spectral reflectance R of the mixture (lambda) from the ratio (KZS) Micromax, can be determined according to equation 3 1. the thickness of the spectral reflectance R (/) is an ideal state (the colorings infinite ), The spectral reflectance R '(λ) that can be measured with a spectrophotometer is calculated according to the following equation 32 (Sanderson's equation).
R' = k1 + (l-k1)(l-k2)1 ...(32) ここで、 係数 , k2 は、 被着色物 (ベース釉) の光学的性質に依存する 値である。 係数 1^ , k , の値は、 被着色物の屈折率 nから次の数式 3 3に従つ て決定することができる c k1 = Ml-J- (33a) R '= k 1 + (lk 1 ) (lk 2 ) 1 ... (32) where the coefficient, k 2 is a value that depends on the optical properties of the object to be colored (base glaze). The values of the coefficients 1 ^, k, can be determined from the refractive index n of the object to be colored according to the following equation 33 : c k 1 = Ml-J- (33a)
\n+ 1  \ n + 1
k =0.68n-0.56 .. - (33b)  k = 0.68n-0.56 ..-(33b)
なお、 ベース釉の屈折率 nは、 例えば約 1. 4である。  The refractive index n of the base glaze is, for example, about 1.4.
数式 3 2によって分光反射率 R' {?, ) が求まると、 混合物の三刺激値 X- , YM , ZH は、 前述した数式 3 0と同様な式に従って算出できる。 When the spectral reflectance R '{?,) Is obtained from Equation 32, the tristimulus values X-, Y M and Z H can be calculated according to the same formula as the above-described formula 30.
ところで、 各着色剤の吸収係数 と散乱係数 S iは誤差を含んでおり、 また、 数式 3 2〜 3 3は理論式ではなく実験式なので、 ステップ S 5 4で得られた 三刺激値の計算値 XM, YM , ZMは誤差を有している。 この計算誤差は、 ステツ プ S 5 2で得られた実測値 X。 , Y。 , Ζ。 からの差である。 ところで、 コン ピュータカラーマッチングによる調合割合の予測も、 数式 1 , 2, 3 0〜 3 3に 従って混合物の三刺激値を求める工程を含んでいる。 そこで、 濃淡限度見本の調 合割合をコンピュータカラ一マッチングで予測する際に、 標準色見本における三 刺激値の計算誤差を考慮しておけば、 その予測精度を向上させることができる。 図 2 0のステップ S 5 5では、 ステップ S 5 2で得られた標準色の三刺激値の 実測俩 X。 , Υ0 , Ζ0 と、 ステップ S 5 4で られた計算値 ΧΜ , ΥΜ , ΖΜ との差分 ΔΧ, ΔΥ, ΔΖを次の数式 3 4に従って求める。 By the way, the absorption coefficient and the scattering coefficient S i of each colorant include an error, and the equations 32 to 33 are empirical rather than theoretical, so the tristimulus values obtained in step S 54 are calculated. The values X M , Y M , and Z M have errors. This calculation error is the measured value X obtained in step S52. , Y. , Ζ. Is the difference from By the way, the prediction of the blending ratio by computer color matching also includes the step of calculating the tristimulus values of the mixture according to the formulas 1, 2, 30 to 33. Therefore, when predicting the mixing ratio of the shading limit sample by computer color matching, if the calculation error of the tristimulus value in the standard color sample is considered, the prediction accuracy can be improved. In step S55 of FIG. 20, an actual measurement 俩 X of the tristimulus value of the standard color obtained in step S52. , Upsilon 0, and Zeta 0, Step S 5 4 out was calculated values Χ Μ, Υ Μ, difference ΔΧ with Ζ Μ, ΔΥ, seek ΔΖ following Equation 3 4.
ΔΧ = Χ0-ΧΜ . . - (34a) ΔΧ = Χ 0 -ΧΜ..-(34a)
ΔΥ = Υ0Μ .. - (34b) ΔΥ = Υ 0Μ ..-(34b)
ΔΖ = Zo - ZM ... (34c) ΔΖ = Zo-Z M ... (34c)
ステップ S 5 6では、 ステップ S 5 3で設定された濃淡限度見木の三刺激値 X τ , Υ.,. , Ζτ を上言 tiの差分 ΔΧ, ΔΥ, 厶 Ζで補正することによって、 コン ピュータカラーマッチングにおける三刺激値の 標値 Xc , Yc , Zc を求め る。 すなわち、 渙淡限度見本に対する三刺激値の目標値 Xc , Yc , Zc は次 の数式 3 5で与えられる。 In step S56, the tristimulus values X τ , Υ.,., Ζ τ set in step S53 are corrected by the differences ΔΧ, ΔΥ, Ζ of the above ti. target value X c tristimulus values in computer color matching, Y c, Ru seek Z c. That is, the target values X c , Y c , and Z c of the tristimulus values for the lysis limit sample are given by the following Expression 35.
Χτ = Χτ-ΔΧ ... (35a) Χτ = τ τ -ΔΧ ... (35a)
ΥΓ = ΥΤ-ΔΥ ... (35b) Υ Γ = Υ Τ -ΔΥ ... (35b)
ΖΓ = ΖΤ-ΔΖ ... (35C) ΖΓ = Ζ Τ -ΔΖ ... (35C)
ステップ S 5 7では、 コンピュータカラーマッチングによって濃淡限度見本の 調合割合を予測する。 このコンピュータカラーマッチングでは、 数式 3 5で与え られる目標値 Xc , γハ , Zc が得られるような調合割合を求めている。 図 2 1は、 ステップ S 5 7におけるコンピュータカラーマッチングの詳細丰順 を示すフローチャートである。 なお、 図 2 1の手順は、 ニュートン一ラプソン法 を用いたコンピュータカラーマッチングを適用したものである。 In step S57, the mixing ratio of the density limit sample is predicted by computer color matching. In this computer color matching, the mixing ratio is determined so that the target values Xc , γc, and Zc given by Expression 35 are obtained. FIG. 21 is a flowchart showing the detailed procedure of the computer color matching in step S57. The procedure in Fig. 21 applies computer color matching using the Newton-Raphson method.
ステップ S 6 】では、 各着色剤 (顔料) の調合率 Ci を微小変化させた時の三 刺激値の変化を以下の手順で計算する。 まず、 図 20のステップ S 5 2で用いた 標準色 ¾本の調合率 Ci から、 〗つの着色剤の調合率のみを AC だけ変えて、 その混合物の 刺激値の計算値 ΧΜ', ΥΜ', ΖΜ 'を求める。 これは、 前述した数 式 1, 2, 3 0〜3 3を用いて、 ステップ S 5 2と同様の手順で行なわれる。 そ して、 ステップ S 5 2で得られていた標準色の三刺激値の計算値 Χ„ , ΥΜ , Ζ Μ との差分 AXCi, AYci, AZCiを以下の数式 3 6に従って算出する。
Figure imgf000041_0001
In step S6], the change in tristimulus value when the mixing ratio Ci of each colorant (pigment) is slightly changed is calculated in the following procedure. First, from the mixing ratio Ci of the standard color used in step S52 in FIG. 20, only the mixing ratio of one colorant is changed by AC, and the calculated stimulus values of the mixture Χ Μ ', Υ Μ ', Ζ Μ '. This is performed in the same manner as in step S52, using the above-described equations 1, 2, 30 to 33. Then, the differences AX Ci , AY ci , and AZ Ci from the calculated values 三, Υ ,, Ζ of the tristimulus values of the standard color obtained in step S 52 are calculated according to the following equation 36. .
Figure imgf000041_0001
ΔΥΟ, = ΥΜ-ΥΜ ...(36b)  ΔΥΟ, = ΥΜ-ΥΜ ... (36b)
ΔΖΟ; = Ζ -ΖΜ . - . (36c) ΔΖΟ;.. = Ζ -Ζ Μ - (36c)
数式 36を用いると、 各着色剤の調合率 Ci のみを微小変化させた時の三刺激 値の変化率が、 次の数式 3 7で与えられる。
Figure imgf000041_0002
Using Equation 36, the change rate of the tristimulus value when only the mixing ratio Ci of each colorant is slightly changed is given by the following Equation 37.
Figure imgf000041_0002
)
^ΔΥο, ...(37b)
Figure imgf000041_0003
濃淡限度色は、 標準色の製造誤差を示すものなので、 両者の三刺激値は互いに 近接した値である。 従って、 数式 3 7の変化率を用いると、 図 20のステップ S 5 6で得られた濃淡限度見本の三刺激値の目標値 Xc , Yc , Zc と、 ステツ プ S 5 4で得られた標準色見本の三刺激値の計算値 XM, YH, ZMとの関係を、 次の数式 38で表わすことができる。
Figure imgf000042_0001
^ ΔΥο, ... ( 37b)
Figure imgf000041_0003
Since the shading limit color indicates a manufacturing error of the standard color, the tristimulus values of both are close to each other. Thus, with the rate of change of Equation 3 7, obtained in step S 5 the target value X c tristimulus values of the resulting shade limit sample at 6, Y c, and Z c, Sutetsu flop S 5 4 in FIG. 20 The relationship between the calculated tristimulus values X M , Y H , and Z M of the obtained standard color samples can be expressed by the following equation (38).
Figure imgf000042_0001
... (38b)
Figure imgf000042_0002
... (38b)
Figure imgf000042_0002
△C , + AC2 +AC3 +ΔΟ) = APVC ... (38d) △ C, + AC 2 + AC 3 + ΔΟ) = APVC ... (38d)
なお、 数式 38では、 着色剤が 4種類であると仮定している。 なお、 各着色剤 の調合率の変化量 の合計値 ΔΡ CVには一定値が代入される。 例えば、 4 種類の着色剤の調合率 Ci の和∑Ci を一定に保った場合には、 ? は0で ある。  Note that Equation 38 assumes that there are four types of colorants. Note that a fixed value is substituted for the total value ΔΡCV of the change amount of the mixing ratio of each colorant. For example, if the sum ∑Ci of the mixing ratios Ci of the four colorants is kept constant,? Is 0.
数式 3 8は、 4つの未知数 ( i = 1〜4) を舍む 4元連立一次方程式な ので、 これを解くことによって、 各未知数 の値を求めることができる (ス テツプ S 62) 。 ステップ S 63では、 ステップ S 62で求められた彼 を 用いて、 濃淡限度見本の各着色剤の調合率 CiTが次の数式 3 9に従って算出され る。 Since Equation 38 is a four-dimensional system of linear equations involving four unknowns (i = 1 to 4), the value of each unknown can be obtained by solving this equation (Step S62). In step S63, using him obtained in step S62 , the mixing ratio C iT of each colorant of the shading limit sample is calculated according to the following Expression 39 .
ClT = Ci + ACi .-.(39) C lT = Ci + ACi .-. (39)
ステップ S 64では、 ステップ S 63で得られた調合率 CiTから、 前述した数 式 1, 2, 30〜 3 3に従って三刺激値の計算値 , Υχ , Z2 を求める。 ス テツプ S 65では、 こうして得られた三刺激値 , Υ: , Z: と、 図 20のス テツプ S 5 6で求めた三刺激値の目標値 Xc , Yc , Zc との色差 が所定 の許容誤差 <5以下であるか否かが判断される。 なお、 色差 は、 L a b表色系 を用いて次の数式 40で与えられる。 In step S 64, from the blending ratio C iT obtained in step S 63, the number mentioned above Equation 1, 2, calculated values of tristimulus values according to. 30 to 3 3, Upsilon chi, seek Z 2. In scan Tetsupu S 65, thus resulting tristimulus values, Upsilon:, Z: the target value X c tristimulus values obtained in scan Tetsupu S 5 6 of FIG. 20, Y c, the color difference between Z c It is determined whether or not the predetermined tolerance is less than or equal to <5. The color difference is given by the following equation 40 using the Lab color system.
ΔΕ = ( - L,)2 + (ac - a !i)2 + (bc - b,!)2 ...(40) ΔΕ = (-L,) 2 + (a c -a! I) 2 + (b c -b ,!) 2 ... (40)
ステップ S 65において、 色差 ΔΕが許容誤差 όよりも小さな場合にはステツ プ S 63で得られた調合率 CiTの値を予測値として採用し、 コンピュータカラー マッチングを終了する。 一方、 色差厶 Eが許容誤差 ό以上の場合には、 ステップ S 66において、 ステップ S 64で得られた三刺激値の計算値 Χ1 , Υ1 , Ζ, を、 ΧΜ , ΥΜ , ΖΜ に代入して、 ステップ S 62〜S 65の処理を繰り返す。 こうしてステップ S 62〜S 6 6を繰り返し実行することによって、 色差 ΔΕの 値が許容誤差 <5よりも小さくなるような調合率 CiTを求めることができる。 In step S65, if the color difference Δ 許 容 is smaller than the allowable error S, The value of the mixing ratio CiT obtained in step S63 is adopted as the predicted value, and the computer color matching is terminated. On the other hand, if the color difference E is equal to or larger than the allowable error ό, the calculated values 三1 , 三1 , Ζ, of the tristimulus values obtained in step S 64 are replaced with Χ Μ , Υ ,, Ζ in step S 66. Μ and the processing of steps S62 to S65 is repeated. By repeatedly executing steps S 62~S 6 6 Thus, the value of the color difference ΔΕ can be obtained compounding ratio C iT as smaller than the allowable error <5.
図 22は、 本発明の第 3実施例において用いた標準色見本の三刺激値と、 濃淡 限度見本の三刺激値を示すテ一ブルである。 サンプル T l, Τ 2, Τ 3は、 それ ぞれ異なる色を有する標準色見本である。 図 2 2には、 各標準色見本について、 図 20のステップ S 5 2で実測された三刺激値 X。 , Υ0 , Ζ,, と、 ステップ S 5 4で計算された三刺激値 ΧΜ, ΥΜ , ΖΜ とが示されており、 また、 ステツ プ S 5 3で設定された濃淡限度見本の三刺激値の設定値 Χτ , Υτ , Ζτ も示 されている。 コンピュータカラーマッチングにおける濃淡限度見本の三刺激値の 目標値 Xc , Yc , .Zc は、 これらの値から、 前述した数式 34および 35に 従って算出される。 FIG. 22 is a table showing tristimulus values of a standard color sample and tristimulus values of a shading limit sample used in the third embodiment of the present invention. Samples T l, Τ 2, Τ 3 are standard color samples having different colors. Fig. 22 shows the tristimulus values X actually measured in step S52 in Fig. 20 for each standard color sample. , Upsilon 0, and Zeta ,,, Step S 5 4 tristimulus values chi Micromax calculated in, Upsilon Micromax, are shown and Zeta Micromax is also shading boundary samples set in Sutetsu flop S 5 3 The set values of the tristimulus values Χ τ , Υ τ , and Ζ τ are also shown. The target values X c , Y c , .Z c of the tristimulus values of the gray scale limit sample in the computer color matching are calculated from these values according to the above-mentioned equations 34 and 35.
図 23は、 第 3実施例における濃淡限度見本の三刺激値の予測結果と、 比較例 の予測結果と、 実調合率とを示したものである。 比較例の調合率は、 濃淡限度見 本の三刺激値の設定値 Χτ , Υτ , Ζτ をそのままコンピュータカラーマッチ ングの目標値とした時に得られる予測値である。 また、 第 3実施例の調合率は、 標準色見本の計算誤差で補正した目標値 Xc , Yc , Zc を用いて得られる予 測値である。 なお、 この第 3実施例では、 予測精度を検証するために、 図 2 3の 右端に示す調合率を用いて濃淡限度見本を実際に製作しておき、 その三刺激値の 実測値を図 22に示す設定値 Χτ , Υτ , Ζτ として用いている。 FIG. 23 shows the prediction results of the tristimulus values of the density limit sample in the third embodiment, the prediction results of the comparative example, and the actual blending ratio. The blending ratio of the comparative example is a predicted value obtained when the set values 三τ , Υ τ , Ζ τ of the tristimulus values of the gray scale limit sample are directly used as the target values of the computer color matching. Also, compounding ratio of the third embodiment is the predicted value obtained using the target value X c were corrected in the calculation error of the standard color swatch, Y c, the Z c. In this third example, in order to verify the prediction accuracy, a gray scale limit sample was actually manufactured using the mixing ratio shown on the right end of FIG. 23, and the measured tristimulus values were compared with those in FIG. Are used as the set values Χ τ , Υ τ , Ζ τ shown in Fig.
図 2 3から解るように、 第 3実施例は比較例に比べてより高い精度で調合率を 予測することが可能であった。 また、 3つのサンプル丁】〜 Τ 3では 4つの顔料 の成分がかなり異なるが、 そのすベての場合において、 第 3実施例の予測精度が 比較例に比べて高いことが解る。 以上のように、 上記第 3実施例では、 瀘淡限度見本に近い三刺激値を有する標 準色見本に関して三刺激摘の計算誤差を求め、 これを用いて濃淡限度見本の三刺 激値の目標値を補正している。 この結果、 濃淡限度見本の調合率を予測する際の 予測精度を向上させることが可能である。 As can be seen from FIG. 23, the third example was able to predict the mixing ratio with higher accuracy than the comparative example. In addition, although the components of the four pigments are significantly different in the three samples, the prediction accuracy of the third embodiment is higher than that of the comparative example in all cases. As described above, in the third embodiment, the calculation error of tristimulation was calculated for a standard color sample having a tristimulus value close to the filtering limit sample, and this was used to calculate the tristimulus value of the shading limit sample. The target value has been corrected. As a result, it is possible to improve the prediction accuracy when predicting the mixing ratio of the shading limit sample.
と記第 3実施例に関しては、 例えば次のような変形も可能である。  With respect to the third embodiment, for example, the following modifications are also possible.
( 1 ) 上記第 3実施例では、 濃淡限度見本に近い三刺激値を有するサンプル (近 接色サンプル) として標準-色見本を使用し、 標準色見本に関する三刺激値の計算 誤差を用いて浪淡限度見本の三刺激値の目標値を補正していた。 しかし、 標準色 見本以外の近接色サンプルを選択し、 その近接色サンプルに関する計算誤差を用 いることも可能である。 例えば、 種々の色を有するサンプルを収集したデータべ一 スを用いて、 濃淡限度見本に近い色を有するサンプルを、 近接色サンプルとして 選択するようにしてもよい。 このデータベースは、 各着色剤の調合率 C i と、 三 刺激値の実測値 (または、 反射率 R ( λ ) ) を少なくとも舍むようにするのが好 ましい。 なお、 データベースを用いて近接色サンプルを検索する場合には、 上述 の数式 4 0で与えられる色差 が、 目標とする混合物 (濃淡限度見本) に最も 近いものを近接色サンプルとすることが好ましい。  (1) In the third embodiment, the standard-color sample is used as a sample having a tristimulus value close to the density limit sample (near-color sample), and the calculation error of the tristimulus value for the standard color sample is used. The target value of the tristimulus value of the light-limit sample was corrected. However, it is also possible to select a nearby color sample other than the standard color sample and use the calculation error for that nearby color sample. For example, using a database in which samples having various colors are collected, a sample having a color close to the shading limit sample may be selected as a nearby color sample. This database should preferably include at least the formula C i of each colorant and the measured tristimulus value (or the reflectance R (λ)). When a close color sample is searched using a database, it is preferable that the closest color sample whose color difference given by Expression 40 is closest to the target mixture (shade limit sample) is used.
なお、 このようなデータベースを用いれば、 漉淡限度見本の調合率をコン ピュータカラーマッチングによって予測する場合に限らず、 任意の三刺激値を有 する目標混合物の調合率を予測する場合に、 その予測精度を向上させることがで きるという利点がある。  It should be noted that using such a database is not limited to the case where the blending rate of the sample of the strain limit is predicted by computer color matching, but is used to predict the blending rate of a target mixture having arbitrary tristimulus values. There is an advantage that the prediction accuracy can be improved.
( 2 ) 上記第 3実施例では、 Χ Υ Ζ表色系を用いていたが、 表色系としては Χ Υ Ζ表色系以外の任意の表色系を使用することができる。 例えば、 L * a * b * 表 色系などを使用することが可能である。 D . 第 4実施例:  (2) Although the third embodiment uses the Χ / Ζ color system, any color system other than the Χ / Ζ color system can be used as the color system. For example, an L * a * b * color system can be used. D. Fourth embodiment:
顔料や染料などの着色剤の着色対象となるもの (被着色物) を所望の色とする には、 その所望の色となるように着色剤を種々調合して調合割合を規定する必要 がある。 この場合、 技術者が試行錯誤して所望の色を呈する着色剤の混合割合を 求めていたが、 近年では、 個々の着色剤の濃度別の光学的データを基礎として、 着色剤の混合割合を予測するいわゆるコンピュータカラーマッチング (以下、 こ れを適宜 C C Mと略称する) が提案されている (特開平 4一 1 8 1 1 2 9, 特公 平 6— 9 8 8 8 0 ) 。 In order to obtain a desired color for an object to be colored by a coloring agent such as a pigment or a dye (object to be colored), it is necessary to formulate the coloring agent in various ways so as to obtain the desired color and to define a mixing ratio. . In this case, the technician determines the mixing ratio of the colorant that exhibits the desired color through trial and error. In recent years, so-called computer color matching (hereinafter abbreviated as CCM) has been proposed in recent years, which predicts the mixing ratio of colorants based on optical data for each colorant concentration. (Japanese Patent Application Laid-Open No. Hei 4-181, 1989, Japanese Patent Publication No. Hei 6-98880).
ところで、 これら公報で提案された C CMでは、 ニュートン 'ラプソン法等の 逐次近似法を用いているため、 その予測した調合割合が負となることがある。 こ のため、 負となった予測調合割合をゼロ若しくは正とするための Π夫がなされて いる。  By the way, in the CCMs proposed in these publications, since the successive approximation method such as the Newton-Raphson method is used, the predicted mixing ratio may be negative. For this reason, efforts have been made to make the negative forecast compounding ratio zero or positive.
上記した従来の C C Mでは、 その被着色物を繊維とし、 一旦着色 (染色) した 繊維の色が所望のものではない場合に、 この染色済みの繊維を追加染色してその 色を所望のものとする。 しかしながら、 陶器やタイルでは、 その色は、 釉に混合 する顔料の調合率 (調合割合) によって決定され、 顔料が調合された釉 (以ド、 釉薬という) の焼成を経て当該色を呈する。 このため、 陶器やタイル等のように 追加染色 (着色) できない被着色物には、 上記の従来の C C Mは適用できない。 また、 追加着色が可能な繊維や追加着色が不可能な陶器, タイルであっても、 共に、 着色剤 (繊維にあっては染料, 陶器やタイルにあっては釉薬) は、 繊維, 陶器等の工業的生産に合わせて多量に繰り返し調合される。 この場合、 着色剤の 調合割合は維持されるが、 調合工程に変動、 例えば温度や調合のタイミング等の ズレがあったりすると、 それ以前に調合した着色剤で得られる色や色見本の色を 再晃できない場合がある。 特に、 陶器やタイルでは、 天然の顔料を用いる都合上、 色の再現の信頼性にやや欠ける。  In the above-mentioned conventional CCM, the object to be colored is a fiber, and if the color of the dyed (dyed) fiber is not the desired one, the dyed fiber is additionally dyed to change the color to the desired one. I do. However, in ceramics and tiles, the color is determined by the mixing ratio (mixing ratio) of the pigment mixed with the glaze, and the color is obtained through firing of the glaze (hereinafter referred to as glaze) in which the pigment is mixed. For this reason, the above-mentioned conventional CCM cannot be applied to an object to be colored that cannot be additionally dyed (colored), such as pottery and tiles. In addition, the colorant (dye for fiber, glaze for pottery and tile) is used for fiber, pottery, etc. Is repeatedly formulated in large quantities according to the industrial production of In this case, the mixing ratio of the colorant is maintained, but if the mixing process fluctuates, for example, if there is a deviation in the temperature, the timing of the mixing, or the like, the color obtained from the colorant previously mixed or the color of the color sample is changed. There is a case that cannot be redone. Especially in ceramics and tiles, the reliability of color reproduction is somewhat lacking due to the use of natural pigments.
このような場合、 追加着色が可能な繊維では、 上記した C C Mにより追加染色 して色修正ができるが、 追加染色するために調合した着色剤と既に調合済みの着 色剤を併用しなければならず、 煩雑である。 この煩雑さを回避するためには、 色 見本の色を得られる着色剤を新たに調合してこの着色剤だけで染色すればよいが、 既に調合済みの着色剤は不要として廃棄するか、 この調合済みの着色剤に新たな 着色剤を追加して再調合する必要がある。 また、 陶器やタイルでは、 追加着色が 不可能な都合上、 調合済みの着色剤の廃棄か再調合を採ることになる。 しかしな がら、 調合済み着色剤の廃棄は無駄であり、 その一方、 着色剤の再調合は技術者 の長年の勘と経験により着色剤を徐々に追加しながら行なうのでやはり煩雑であつ た。 In such a case, for fibers that can be additionally colored, the color can be corrected by additional dyeing using the above-mentioned CCM.However, in order to perform additional dyeing, the colorant that has been prepared must be used in combination with the already prepared colorant. It is complicated. In order to avoid this complication, a colorant that can obtain a color sample color may be newly prepared and dyed with only this colorant. However, the already prepared colorant is discarded as unnecessary or discarded. It is necessary to add a new colorant to the prepared colorant and re-formulate. Also, for pottery and tiles, it is not possible to add additional color, so the prepared colorant must be discarded or remixed. But However, disposing of the prepared colorant is useless, while re-formulation of the colorant is also cumbersome because the colorant is gradually added based on the intuition and experience of engineers.
以下に説明する第 4実施例は、 調合済み着色剤の有効利用を図りつつ、 再調合 を簡略化することを冃的としたものである。  The fourth embodiment described below aims at simplifying re-mixing while effectively utilizing the mixed colorant.
図 2 4は、 実施例のコンピュータカラーマッチング方法を実施するための装置 を示すブロック図である。 この装置は, 上述した第 1〜第 3実施例を実現する装 置としても使用することが可能である。 演算装置 4 0は汎用コンピュータであり, 図示しない C P Uがソフトゥヱァプログラムを実行することによって, この発明 によるコンピュータカラーマツチングの各工程および各手段を実現する。  FIG. 24 is a block diagram illustrating an apparatus for performing the computer color matching method according to the embodiment. This device can also be used as a device that implements the first to third embodiments described above. The arithmetic unit 40 is a general-purpose computer, and realizes each step and each means of computer color matching according to the present invention by causing a CPU (not shown) to execute a software program.
図 2 5は、 この第 4実施例における処理の全体手順を示すフ口一チヤートであ る。 なお、 この第 4実施例で対象とする調合物は、 陶磁器の素地の表面を覆うた めの釉である。 即ち、 顔料を入れないベース釉 (基礎釉) が被着色物であり、 こ のベース釉に顔料を添加した釉がコンピュータカラ一マツチングの対象となる調 合物である。 また、 その表色系は、 X Y Z表色系とするが、 これ以外の表色系、 例えば L * a * b *表色系を採用してもよいことは勿論である。  FIG. 25 is a flowchart showing the entire procedure of the process in the fourth embodiment. The composition targeted in the fourth embodiment is a glaze for covering the surface of a ceramic body. In other words, the base glaze without pigment (base glaze) is the material to be colored, and the glaze with the pigment added to this base glaze is the target for computer color matching. Although the color system is the XYZ color system, it goes without saying that other color systems, for example, the L * a * b * color system may be adopted.
図 2 4に示すように、 第 4実施例では、 C CMに関する後述する稀々の演算を 実行する演算装置 4 0と、 データ入力を行なうキーボード, マウス等の入力機器 4 2と、 演算の状態や後述する合否判定の結果を表示する表示機器 4 4と、 この 合否判定の結果や種々の演算式等を記憶する記憶装置 4 6と、 C CMに必要なデー タとしての分光反射率を取得するための分光光度計 4 8とを備える。 そして、 こ の演算装置 4 0では、 以下の調合処理が行なわれる。  As shown in FIG. 24, in the fourth embodiment, an arithmetic unit 40 for executing a rare operation, which will be described later, relating to the CCM, an input device 42 such as a keyboard and a mouse for inputting data, and a state of the operation And a display device 44 for displaying the results of pass / fail judgment described later, a storage device 46 for storing the results of the pass / fail judgment and various arithmetic expressions, etc., and acquiring a spectral reflectance as data necessary for CCM. And a spectrophotometer 48. Then, in the arithmetic unit 40, the following blending process is performed.
まず、 ステップ S 7 1では、 目標色を呈する釉 (見本釉) を準備する。 この場 合の H標色は、 調合済みの釉で呈される色であるので、 この釉 (見本釉) におけ る顔料の調合割合は既知である。 続くステップ S 7 2では、 この冃標色見本釉を 分光光度計 4 8で測色し、 X Y Z表色系での色評価値である三刺激値 (実測値) Χτ , Υ τ , Ζτ を求める。 この三刺激値は、 分光光度計 4 8での測色により 得られた目標色見本釉の分光反射率 R ' ( - ) から、 次の数式 4 2に従って算出 される。 なお、 算出された三刺激値は、 表示機器 4 4に目標見本色とともに表示 され、 後述の処理に用いるために記憶装置 4 6に記憶される。 また、 以下に記す 三刺激値等の演算結果は、 その都度に、 記憶装置 4 6に記憶される。
Figure imgf000047_0001
First, in step S71, a glaze (sample glaze) exhibiting a target color is prepared. Since the H standard color in this case is the color presented by the prepared glaze, the blending ratio of the pigment in this glaze (sample glaze) is known. In the following step S72, the 冃 standard color sample glaze is measured with a spectrophotometer 48, and tristimulus values (actually measured values) τ τ , , τ , Ζ τ which are color evaluation values in the XYZ color system Ask for. The tristimulus value is calculated from the spectral reflectance R ′ (-) of the target color sample glaze obtained by colorimetry with a spectrophotometer 48 according to the following equation 42. Is done. The calculated tristimulus value is displayed on the display device 44 together with the target sample color, and is stored in the storage device 46 for use in processing described later. The calculation results of the tristimulus values and the like described below are stored in the storage device 46 each time.
Figure imgf000047_0001
YT = K j S( )y( ) R'( )d ... (42b) Y T = K j S () y () R '() d ... (42b)
ZT = K |S( )z( ) R'(X)dX ... (42c) Z T = K | S () z () R '(X) dX ... (42c)
K= 100/ jS(X)y( )dX ... (42d) ここで、 S (;. ) は標準光の分光分布、 X (λ ) , y U) , z (/) (数式 中ではバー付きである。 以下同じ) は等色隱であり、 い" mも既知の^ ある。 なお、 式中に記した ( ) は、 分光反射率, 分光分布, 等色関数がいずれも波長 λに依存していることを表わす。 K = 100 / jS (X) y () dX ... (42d) where S (;.) Is the spectral distribution of standard light, X (λ), y U), z (/) (where (The same applies to the bars.) The same is true for color matching, and “m is also known. The parentheses in the formula indicate that the spectral reflectance, spectral distribution, and color matching function are all wavelengths λ. Indicates that it is dependent on
ステップ S 7 3では、 上記の見本釉が呈する目標色と近似する色を呈すると想 定した調合割合でそれぞれの顔料を添加し、 第】回目の試作釉を調合し、 この試 作釉についてステップ S 7 2と同様にして測色する。 これにより、 この第 1回目 試作釉についての三刺激値 (実測値) , , Z, を求める。 この第】回目 試作において、 上記の調合割合での顔料の調合は技術者によってなされるものの、 その値は任意性のある既知の値であり、 従来のように試行錯誤して繰り返し調合 する必要はなく、 調合割合を定めるに当たって特段の経験や勘も要しない。 この 場合の第 1回目試作釉についての三刺激値並びに顔料調合量 (調合割合) も、 表 示機器 4 4に表示されると共に、 記憶装置 4 6に記憶される。 また、 この際の顔 料調合量は、 入力機器 4 2から入力される。  In step S73, each pigment is added at a blending ratio assumed to exhibit a color similar to the target color exhibited by the sample glaze, and the second trial glaze is blended. The color is measured in the same manner as in S72. The tristimulus values (actually measured values),, and Z for the first prototype glaze are obtained. In this second trial, the pigments are blended at the above blending ratios by a technician, but the values are arbitrary values that are known and need to be repeatedly blended by trial and error as in the past. No special experience or intuition is required to determine the mix ratio. In this case, the tristimulus value and the pigment preparation amount (preparation ratio) for the first prototype glaze are also displayed on the display device 44 and stored in the storage device 46. In addition, the amount of the mixture at this time is input from the input device 42.
続くステップ S 7 4では、 ステップ S 7 1で作成した目標色見本釉の呈する色 とステップ S 7 3における第 1回目試作釉の呈する色との色差 ΔΕ* (JIS Z 8"730)が所定範囲内に納まるか否か、 即ち色差 ΔΕ* の合否判定を下す。 この 場合の色差 ΔΕ* の許容値は、 目標色見本の呈する色と第 1回目試作釉の呈す る色との違いが一見しては判別できない程度の値、 例えば 0. 3〜0. 5程度の 値が予め人力機器 4 2から設定される。 なお、 色差 ΔΕ* の許容値を 0. 3〜 0. 5以外の値とすることもできることは勿論である。 In the following step S74, the color difference ΔΕ * (JIS Z 8 "730) between the color of the target color sample glaze created in step S71 and the color of the first prototype glaze in step S73 is set within a predetermined range. In this case, the allowable value of the color difference ΔΕ * is determined by the difference between the color of the target color sample and the color of the first prototype glaze. Value that cannot be distinguished by The value is set in advance from the manpower device 42. Of course, the allowable value of the color difference ΔΕ * can be set to a value other than 0.3 to 0.5.
このステップ S 7 4で色差 ΔΕ* が 0. 3〜0. 5以内であると合格判定す れば、 ステップ S 7 3での第 1回目試作の釉で目標色見本の色を再 できるので、 それ以上の調合処理は不要であるとして総ての処理を終了する。 つまり、 第 1回 目試作の際の調合割合で調合した新たな釉は、 目標色見本釉とほぼ同じ色を呈す る。  If it is judged that the color difference ΔΕ * is within 0.3 to 0.5 in step S74, the color of the target color sample can be re-used with the glaze of the first prototype in step S73. It is determined that no further compounding processing is necessary, and all the processing ends. In other words, the new glaze blended at the blending ratio at the time of the first trial production has almost the same color as the target color sample glaze.
一方、 ステップ S 7 4で色差厶 Ε* が 0. 3〜0. 5以内に納まらないとし て不合格判定した場合には、 続くステップ S 7 5で、 ステップ S 7 2で求めた目 標色見木の三刺激値 Χτ , Υτ , Ζτ とステップ S 7 3で求めた第 1回目試作 釉の三刺激値 X Υ, , とから、 次の数式 4 3に従って三刺激値の差 ΔΧ, ΔΥ, ΔΖを算出する。 この三刺激値の差 ΔΧ, ΔΥ, ΔΖは、 目標色見木の呈 する色と第 1回 Η試作釉の呈する色との色差△ Ε * を反映した値となる。 On the other hand, if the color difference Ε * does not fall within the range of 0.3 to 0.5 in step S74, the rejection is determined, and in step S75, the target color obtained in step S72 is determined. From the tristimulus values 見τ , Υ τ , Ζ τ of the miki and the tristimulus values X Υ,, of the first prototype glaze obtained in step S 73, the difference between the tristimulus values according to the following equation 43 is ΔΧ , ΔΥ, ΔΖ. The differences ΔΧ, ΔΥ, ΔΖ of the tristimulus values reflect the color difference △ * between the color presented by the target swatch and the color exhibited by the first prototype glaze.
ΔΧ = ΧΤ-Χ) ... (43a) ΔΧ = Χ Τ -Χ) ... (43a)
ΔΥ =ΥΤ- Υ, ... (43b) ΔΥ = Υ Τ -Υ, ... (43b)
ΔΖ = ΖΤ-Ζ】 ,"(43 ΔΖ = Ζ Τ -Ζ], "(43
ステップ S 7 6では、 ステップ S 7 3で行なった第 1回目試作釉の調合時の各 顔料 (調合割合既知) に微量だけその顔料を追加調合した場合の三刺激値の変化 率 (撖係数) を、 C CMの手法を用いて、 以下の手順に従って求める。 なお、 各 顔料を追加調合する際の微量増加調合量も入力機器 4 2から入力される。  In step S76, the rate of change in tristimulus value (撖 coefficient) when a small amount of each pigment (mixing ratio is known) is added to each pigment (mixing ratio is known) at the time of the first trial glaze blending performed in step S73 Is calculated using the CCM method according to the following procedure. It should be noted that a slight increase in the amount of each additional pigment to be prepared is also input from the input device 42.
まず、 このステップ S 7 6の最初のステップ S 9 1では、 図 2 6のフ口一 チャートに示すように、 第 1回目試作釉の呈する色の三刺激値を第 1回目試作時 の各顔料の既知の調合割合から C C Mの手法で演算する。 この場合の三刺激値 (演算値) X1/E , Y1/E , Z1/E は、 次の数式 44で表わされる。
Figure imgf000048_0001
First, in the first step S91 of this step S76, as shown in Fig. 26, the tristimulus value of the color exhibited by the first prototype glaze was calculated for each pigment at the time of the first prototype. Calculate using the CCM method from the known blending ratio. The tristimulus values (calculated values) X 1 / E , Y 1 / E , and Z 1 / E in this case are represented by the following Equation 44.
Figure imgf000048_0001
Y1;E = K ) S( ) y( )R( )d ... (44b) Z1;F: = K j ζ(λ) R (λ) d λ ... (44c) Y 1; E = K) S () y () R () d ... (44b) Z 1; F : = K j ζ (λ) R (λ) d λ ... (44c)
K= 100/ (S( )y( )dX ... (44d) この数式 4 4にあっても、 上記した数式 4 2と同様、 S (A ) は標準光の分光 分布 であり、 X ( , y (λ) , ζ (λ) は等色関数 (既知) である。 し力 し、 数式 44における分光反射率 R U) は、 三刺激値 Χ1/Ε . Υ1/Ε , Ζ1/Ε を顔料の特性から算出する都合上、 やはり顔料の特性を用いて以下のようにして 算出する。 そして、 この算出した分光反射率 R ( J を用いて、 上記の数式 4 4 から三刺激値 X1/E , Y1/E , Z1/E を算出する。 K = 100 / (S () y () dX ... (44d) Even in the equation (44), as in the above equation (42), S (A) is the spectral distribution of the standard light, and X ( , Y (λ) and ζ (λ) are color-matching functions (known), and the spectral reflectance RU in Equation 44 is the tristimulus value Χ 1 / Ε . Ε 1 / Ε , Ζ 1 / For the sake of convenience in calculating Ε from the pigment properties, it is also calculated using the properties of the pigment as follows: Then, using the calculated spectral reflectance R (J, the tristimulus value is obtained from the above equation 44. Calculate X1 / E , Y1 / E , Z1 / E.
被着色物と着色剤の吸収係数 {?.) と散乱係数 Si ( .) とは、 以下の数式 4 5で表わされるダンカン (Duncan) の式と、 数式 4 6で表わされるクベルカ —ムンク (Kubelka-Munk) の混色理論による式の関係にあり、 これら数式に 基づいて、 任意の調合物の分光反射率 R U) を C CMにより求めることができ る。
Figure imgf000049_0001
The absorption coefficient {?.) And the scattering coefficient Si (.) Of the coloring matter and the coloring agent are expressed by the following Duncan's equation represented by the following equation 45 and Kubelka-munk represented by the equation 46. -Munk), the spectral reflectance RU) of any formulation can be determined by CCM based on these equations.
Figure imgf000049_0001
SM= ∑ Q S, ...(45b) SM = ∑ Q S, ... (45b)
i=l ここで、 KM, SM は調合物 (釉) の吸収係数と散乱係数、 Ki , Si は i番 目の成分 (顔料) の吸収係数と散乱係数、 は i番目の成分 (顔料) の調合割 合である。 … この数式 4 6を変形すると、 分光反射率 R {1) は、 次の数式 4 7で与えられ る。 i = l where, K M, S M is the absorption coefficient and scattering coefficient of the formulation (glazed), Ki, Si is the absorption coefficient and scattering coefficient of the i th component (pigment), the i th component (pigment ). … By transforming this formula 46, the spectral reflectance R {1) is given by the following formula 47.
(47a)(47a)
Figure imgf000049_0002
(47b)
Figure imgf000049_0002
(47b)
Figure imgf000050_0001
ここで、 調合物の吸収係数と散乱係数の比 (KZS) は、 この数式 47で示さ れているように、 各顔料の吸収係数 Ki (A) と散乱係数 Si (; .) と調合率 Ci で規定される。 従って、 分光反射率 R (λ ) は、 この比 (KZS) から算出され、 求めることができる。 なお、 Kw , Sw は白色成分 (白顔料) の吸収係数と散 乱係数、 Cwは白色成分の調合割合である。
Figure imgf000050_0001
Here, the ratio between the absorption coefficient and the scattering coefficient (KZS) of the preparation is represented by the absorption coefficient Ki (A), the scattering coefficient Si (;.) And the mixing ratio Ci of each pigment, as shown in Equation 47. Is defined by Therefore, the spectral reflectance R (λ) can be calculated and calculated from this ratio (KZS). K w and S w are the absorption coefficient and scattering coefficient of the white component (white pigment), and C w is the blending ratio of the white component.
続くステップ S 92では、 各顔料 (顔料 1, 2, 3) について、 第 1回目試作 釉の調合割合をそれぞれ微量の調合量だけ侗別に増加させた釉の呈する色につい ての三刺激値 (演算値) X1/1/E , Y1/1/E , Z1/1/E を、 第 1回目試作時の各顔料 の既知の調合割合並びに増加させた顔料の既知の調合割合から C CMの手法で演 算する。 より詳しく説明すると、 まず、 第〗回目試作釉に顔料 1を微量の調合量 (0. 1 *C, ) だけ加え他の顔料 2 3および白顔料は第 1回目試作釉と同一 調合率の釉の呈する色についての三刺激値を演算する。 この場合にあっても、 上 記した数式 44 数式 47が用いられ、 数式 47における は +0. 1In the following step S92, for each pigment (pigments 1, 2, and 3), the tristimulus value (calculation) for the color of the glaze, in which the proportion of the first prototype glaze was separately increased by a small amount, was calculated. Value) X1 / 1 / E , Y1 / 1 / E , Z1 / 1 / E were calculated from the known blending ratio of each pigment at the time of the first trial production and the known blending ratio of the increased pigment. Calculate using the method described above. To explain in more detail, first, add a small amount of pigment 1 (0.1 * C,) to the first trial glaze, and add other pigments 23 and white pigment to the glaze with the same blending ratio as the first trial glaze. The tristimulus value for the color presented by is calculated. Even in this case, the above Equation 44 Equation 47 is used, and in Equation 47, +0.1
* ) となり、 C2 , C3 , Cwはそのままである。 そして、 このように計算 された三刺激値 X1/1/E Y1/1/E , z1/1/E は、 顔料 1の微量増加調合に起因した 三刺激値である。 *), And C 2 , C 3 , and C w remain as they are. Then, thus calculated tristimulus values X 1/1 / E Y 1 /1 / E, z 1/1 / E is a tristimulus value due to trace amounts increase preparation of pigment 1.
同様に、 顔料 2を微量の調合量 (0. 1 *C2 ) だけ加え他の顔料は同一調合 率の釉の呈する色についての三刺激値 X1/2/E Y1/2/E Z1/2/E と、 顔料 3を微 量の調合量 (0. 】 *C3 ) だけ加え他の顔料は同一調合率の釉の呈する色につ いての三刺激値 X1/3/E Y1/3/E Z1/3/E とを演算する。 なお、 白顔料について もその三刺激値 X1/w/E Y1/w/E z1/w/E が同様に演算される。 そして、 各顔料 (顔料 1 2 3および白顔料) をそれぞれ微量増加して調合したことによる三 刺激愤の変化量と各顔料の微量増加量とから、 次の数式 48に従って三刺激俄の 変化率 (微係数) を算出する。 その後は図 25のステップ S 77に進む。 Similarly, Pigment 2 is added in only a small amount (0.1 * C 2 ), and the other pigments are tristimulus values X 1/2 / E Y 1/2 / E Z for the color of the glaze of the same mixing ratio. 1/2 / E , and pigment 3 with only a small amount (0.) * C 3 ), and the other pigments have the same stimulus tristimulus value for the color of the glaze X 1/3 / E Calculate Y 1/3 / E Z 1/3 / E. The tristimulus value X 1 / w / E Y 1 / w / E z 1 / w / E is similarly calculated for the white pigment. Then, based on the amount of change in tristimulus に よ る due to each pigment (pigment 123 and white pigment) being slightly increased and mixed and the amount of slight increase in each pigment, the rate of change of tristimulus is calculated according to the following equation 48. (Derivative coefficient) is calculated. Thereafter, the process proceeds to step S77 in FIG.
顔料 1 微量増加調合: d6fcvJ Pigment 1 Minor increase formulation: d6fcvJ
Figure imgf000051_0001
Figure imgf000051_0001
ステップ S 7 7では、 目標色見本釉と第 1回 試作釉の測色を経てステップ S 7 5で求めた三刺激値の差 ΔΧ, ΔΥ, ΔΖ (実測値の差) を補正するために必 要な各顔料の追加増量調合量 (顔料 1は , 顔料 2は AC2 , 顔料 3は AC 3 , 白顔料は ACW) を、 次の数式 4 9と数式 5 0を用いて算出する。 In step S77, it is necessary to correct the differences Δ 三, ΔΥ, ΔΖ (actual measurement values) of the tristimulus values obtained in step S75 through colorimetry of the target color sample glaze and the first prototype glaze. The required additional compounding amount of each pigment (Pigment 1, AC 2 for Pigment 2 , AC 3 for Pigment 3 , AC W for white pigment) is calculated using the following Equations 49 and 50.
ΔΧ ACW ...(49a)ΔΧ AC W ... (49a)
Figure imgf000052_0001
Figure imgf000052_0001
ΔΥ =—— ACi +—— ACつ +—— AC +—— ACW ... (49b) ΔΥ = —— ACi + —— AC ++ — AC + —— AC W ... (49b)
dCi ec2 ac3 dcw dCi ec 2 ac 3 dc w
ΔΖ = ACW ... (49c)ΔΖ = AC W ... (49c)
Figure imgf000052_0002
Figure imgf000052_0002
F= AiAC,+ A2AC2 + A3AC3 + AVACW ... (50) F = AiAC, + A 2 AC 2 + A 3 AC 3 + A V AC W ... (50)
数式 4 9は、 三刺激値の差 ΔΧ, ΔΥ, 厶 Ζ (実測値の差) を補正するための ものであり、 各顔料の追加増量調合量 (ACi , AC2 , AC3 , ACW) を変 数とする。 また、 数式 5 0は、 各顔料を変動させた場合の費用関数であり、 各顔 料の追加増量調合量 (Δ(^ , AC2 , 厶 C3 , ACW) をやはり変数とする。 従って、 数式 4 9, 5 0を 4つの上記変数について解くことでこれら変数、 即ち 各顔料の追加増量調合量 (厶(^ , AC2 , AC3 , ACW) が定まる。 なお、 数式 5 0における は、 顔料 i を単位量変動させるのに必要な费用である。 この場合、 ステップ S 7 7では、 厶 が厶 ≥ 0を満たし、 ACWが ACW ≥ 0を満たし、 且つ、 数式 5 0で表わされる費用関数 Fを最小とする Δ<^ 並び に ACWを線形計画法を用いて解く。 これにより、 並びに ACWが負の値 として算出されることはない。 Equation 4-9, the difference between the tristimulus values ΔΧ, ΔΥ, provided for correcting厶Ζ (difference measured value), additional bulking formulation amount of each pigment (ACi, AC 2, AC 3 , AC W) Is a variable. Equation 50 is a cost function when each pigment is varied, and the amount of addition of each pigment (Δ (^, AC 2 , m C 3 , AC W ) is also a variable. , equation 4 9, 5 0 these variables by solving the four above variables, i.e. additional increase formulation amount of each pigment (厶 (^, AC 2, AC 3 , AC W) is determined. Here, in equation 5 0 is an expense required to a unit variation amount of the pigment i. in this case, in step S 7 7,厶satisfies厶≥ 0, AC W satisfies AC W ≥ 0, and, in equation 5 0 the AC W a cost function F to delta <^ sequence that minimizes represented solved using linear programming. Accordingly, and never AC W is calculated as a negative value.
続くステップ S 7 8では、 各顔料 (白顔料を含む) の調合割合を、 求めた追加 増量調合量を加味して算出し、 追加増量調合した釉についての物性値を求め更新 する。 より具体的には、 この調合割合と各顔料についての吸収係数および散乱係 数等を用い、 上記の数式 4 5〜数式 47に従って、 分光反射率 R (; -) を求める。 そして、 その後のステップ S 7 9では、 求めた分光反射率 R (λ ) を数式 4 4に 代入して、 各顔抖を追加増量調合した釉の呈する色についての三刺激値 (演算値)In the following step S78, the blending ratio of each pigment (including white pigment) is calculated by taking into account the determined amount of added bulk compounding, and the physical property values of the additionally blended glaze are calculated and updated. I do. More specifically, the spectral reflectance R (;-) is obtained from the formulas 45 to 47 using the blending ratio, the absorption coefficient and the scattering coefficient of each pigment, and the like. Then, in the subsequent step S79, the calculated spectral reflectance R (λ) is substituted into Equation 44, and the tristimulus value (calculated value) for the color of the glaze that is obtained by adding and increasing each face bridge is added.
XEND/E ^B D/E' ZBOTJ/E¾ "演算 る。 XEND / E ^ B D / E 'ZBOTJ / E¾ "Calculate.
ステップ S 8 0では、 この追加増量調合後の三刺激値 XEND/E, YEND/E, ZEND/E. と目標色見木釉についての三刺激値 Χτ , Υτ Ζτ を用い(JIS Z 8730)、 その色差 ΔΕ* が所定範囲内に納まるか否かの合否判定を再度下す。 このステツ プ S 8 0で色差 ΔΕ* が 0. 3 0. 5以内であると合格判定すれば、 それま でのステップで求めた補正量で調合した釉であれば目標色見本の色を再現できる ことになる。 よって、 この場合は、 調合処理が完了したとして処理を終了する。 つまり, 求めた各顔料についての補正量 (追加増量調合量) を加味した調合割合 力 調合済みの釉を再調合する際の最終的な調合割合として採用される。 より具 体的に説明すると、 調合工程の変動等により色に変調を来たした調合済み釉は、 その調合割合が既知であるので第 4実施例における第 1回目試作釉に相当するの で、 この色に変調を来たした調合済み釉に上記の補正量に従った配合割合で各顔 料を追加して再調合すれば、 目標色を呈する釉を再調合できることになる。 In step S80, the tristimulus values X END / E , Y END / E , Z END / E. And the tristimulus values 目標τ , Υ τ Ζ τ for the target swatches are used. (JIS Z 8730), the pass / fail judgment is again made as to whether or not the color difference Δ 納 * falls within a predetermined range. If it is judged that the color difference ΔΕ * is within 0.30.5 in this step S80, if the glaze prepared with the correction amount obtained in the previous steps reproduces the color of the target color sample You can do it. Therefore, in this case, the process is terminated assuming that the blending process is completed. In other words, the blending ratio that takes into account the correction amount (additional blending amount) obtained for each pigment is used as the final blending ratio when the glaze that has been blended is reblended. More specifically, the blended glaze whose color has been modulated due to variations in the blending process, etc., corresponds to the first prototype glaze in the fourth embodiment because the blending ratio is known, If each pigment is added to the blended glaze whose color has been modulated and added at the blending ratio according to the correction amount described above, the glaze exhibiting the target color can be blended again.
一方、 ステップ S 8 0で色差 ΔΕ* が 0. 3 0. 5以内に納まらないとし て不合格判定した場合には、 微係数計算, 補正量計算のための三刺激値をそれま での値からシフトするとともに、 各顔料の微量増加量を変更する。 具体的に説明 すると、 数式 4 3における三刺激値 , Υ, , ZX に、 ステップ S 7 9で求め た三刺激値 XBND/E YBND/E, をシフトする。 これより、 このシフト後の三 刺激値 (X /E, YEND/E, Z /E) から、 ステップ S 7 5では目標色見本との三 刺激偾の差 (厶 X ΔΥ, ΔΖ) が新たに算出される。 On the other hand, if the color difference ΔΕ * is determined not to be within 0.30.5 in step S80, and the rejection is determined, the tristimulus values for differential coefficient calculation and correction amount calculation are set to the previous values. And change the slight increase of each pigment. More specifically, the tristimulus value X BND / E Y BND / E , which is obtained in step S79 , is shifted to the tristimulus values, Υ,, and Z X in Expression 43. From the tristimulus values (X / E , YEND / E , Z / E ) after this shift, the difference (mmXΔΥ, ΔΖ) between the tristimulus and the target color sample is newly obtained in step S75. Is calculated.
また、 各顔料 (顔料 1 2, 3) について、 ステップ S 7 6で考慮した微量增 加量を、 それまでの調合量 (この場合には、 顔料 1は Ci + 0. 1 氺 ) に 0. 1を乗じた微量増加量 ( ( ^ + 0. 1 ) * 0. 1 ) に変更する。 これに より、 各顔料は、 ステップ S 8 0で不合格とされた場合よりそれぞれ 0. 〗だけ 微量増量されるので、 ステップ S 7 6で、 この微 ftif量をカ喊した調合割合から、 各顔料カ^量 ±tftされたことによる≡ί【撤値が上記した数式 4 4〜数式 4 7に従つ て求められる。 そして、 各顔料が微量増加されたことによるこの新たな三刺激値 が数式 4 8における三刺激値 X1/1/E, Υ 1/1/Ε , Z1/1/E 等に替わって用いられて 改めて微係数が算出される。 その後は既述したように各顔料の追加増量調合量 (補正量) が求められ、 ステップ S 8 0で合格判定されるまで、 上記の処理が繰 り返される。 In addition, for each pigment (pigments 1 2 and 3), add the microaddition amount considered in step S76 to the previous formulation amount (in this case, pigment 1 is Ci + 0.1 氺). Change to a small increase ((^ + 0.1) * 0.1) multiplied by 1. This allows each pigment to be 0. 0 each less than if it were rejected in step S80. Since the amount is increased by a small amount, in step S76, the amount of each pigment is ± tft, based on the proportion of the fine ftif that has been reduced. It is required according to. Then, this new tristimulus value due to the slight increase of each pigment is used in place of the tristimulus values X 1/1 / E , Υ 1/1 / Ε , Z 1/1 / E, etc. in Equation 48 . Then, the derivative is calculated again. After that, as described above, the amount of additional compounding (correction amount) of each pigment is determined, and the above-described processing is repeated until a pass is determined in step S80.
以上説明したように第 4実施例のコンピュータカラーマッチング方法では、 そ れまで存在していた釉 (目標色見本釉) に対して、 その呈する色がある程度近似 した色を呈するような調合割合で第 1回目試作釉を調合し、 その後、 この第 1问 目試作釉に、 標色見本釉の呈する色と所定範囲で一致するような調合割合で各 顔料を追加させた場合の追加増量調合量を求める。 このため、 顔料の除去を意味 する負の調合量を C CMにより求めることがない。 よって、 第 4実施例のコン ピュータカラーマッチング方法によれば、 顔料が調合済みの釉 (着色剤) の廃棄 が不要となるために既存の釉を有効に利用することができる。 また、 既存の釉の 利用に当たり、 技術者が関与する工程はステップ S 7 3における一回限りの試作 釉の調合であり、 この際に、 技術者の長年の勘や経験を要しないので、 釉の再調 合を簡略化することができる。  As described above, in the computer color matching method of the fourth embodiment, the glaze (target color sample glaze) that has existed up to that point has a blending ratio such that the color presents a color that is somewhat similar. After blending the first trial glaze, and then adding each pigment to this first trial glaze at a blending ratio that matches the color of the standard sample glaze within a predetermined range, Ask. For this reason, the amount of negative compounding that means the removal of pigment is not determined by CCM. Therefore, according to the computer color matching method of the fourth embodiment, it is not necessary to dispose of the glaze (colorant) in which the pigment has been prepared, so that the existing glaze can be effectively used. In addition, when using the existing glaze, the process involving the engineer is the one-time trial production of the glaze in step S73. At this time, since the engineer's many years of intuition and experience are not required, the glaze is used. Can be simplified.
更に、 第 4実施例のコンピュータカラ一マッチング方法では、 釉を再調合する ための各顔料の補正量の算出の際に、 各顔料の補正量 (A C±並びに A CW) を、 数式 5 0で表わされる費用関数 Fを用いた線形計画法の丰法で求め、 各顔料の補 正に要する費用が最小となるようにした。 このため、 第 4実施例のコンピュータ カラーマツチング方法によれば、 上記した既存の袖の有効利用と再調合に簡略化 に加え、 コスト低減をも図ることができる。 Further, in the computer color matching method of the fourth embodiment, when calculating the correction amount of each pigment for re-mixing the glaze, the correction amount (AC ± and AC W ) of each pigment is calculated by Equation 50. The cost was calculated by the linear programming method using the expressed cost function F so that the cost required to correct each pigment was minimized. For this reason, according to the computer color matching method of the fourth embodiment, in addition to simplifying the above-mentioned effective use and re-mixing of the existing sleeves, the cost can be reduced.
次に、 上記した第 4実施例のコンピュータカラーマッチング方法を行なった際 の各ステップでの処理により得られる三刺激値ゃ微係数について、 図 2 7を用い て説明する。  Next, the tristimulus value ゃ differential coefficient obtained by the processing in each step when the computer color matching method of the fourth embodiment is performed will be described with reference to FIG.
図 2 7は、 ステップ S 7 2で取得した目標色見本釉についての三刺激値 (色値) とステップ S 7 3で取得した第 1回目試作釉についての三刺激値の対比、 並びに 目標色見本釉, 第 1回目試作釉における各顔料の調合率 (調合率) を表わす。 そ して、 図中の Δ Ε * は、 目標色見本釉と第 1回目試作釉との間の色差であり、 この値に基づいてステップ S 7 4での合否判定が下される。 Figure 27 shows tristimulus values (color values) for the target color sample glaze obtained in step S72. And the tristimulus values of the first trial glaze obtained in step S73 and the blending ratio (mixing ratio) of each pigment in the target color sample glaze and the first trial glaze. Δ Ε * in the figure is a color difference between the target color sample glaze and the first prototype glaze, and a pass / fail judgment is made in step S74 based on this value.
図 2 8は、 図 2 7に掲げる第 1回目試作釉に各顔料をそれぞれ微量だけずつ追 加調合した場合の三刺激値の変化率 (微係数) を示しており、 上記の数式 4 8か ら演算される。 この際、 顔料 1 (赤顔料) の微量増量の場合における微係数算出 には、 演算した三刺激値 X 1/1/E, Y 1/1/E, Z1/1/Eが用いられ、 顔料 2 (黄顔料) では三刺激値 X 1/2/E , Y 1/2/E , Z1/2 /E 力、 顔料 3 (青顔料) では三刺激値 X 1/3/E , Y "3/E , z1/3/E 力;、 白顔料では三刺激値 X 1/w/E , Y 1/H/E , ζ 1 ί/Ε が 用いられる。 Figure 28 shows the rate of change (differential coefficient) of tristimulus values when each pigment was added to the first prototype glaze shown in Figure 27 in a very small amount. It is calculated from At this time, the calculated tristimulus values X 1/1 / E , Y 1/1 / E , and Z 1/1 / E are used to calculate the differential coefficient when the amount of pigment 1 (red pigment) is slightly increased. In pigment 2 (yellow pigment) tristimulus values X 1/2 / E, Y 1/2 / E, Z 1/2 / E power, the pigment 3 (blue pigment) the tristimulus values X 1/3 / E, Y 3 / E , z 1/3 / E force; For white pigments, tristimulus values X 1 / w / E , Y 1 / H / E , {1} / ί are used.
この図 2 9は、 第 4実施例によるコンピュータカラーマッチング方法による結 果を示しており、 各顔料について定めた最終的な調合割合で調合した釉と目標色 見本釉とを、 その調合率と三刺激値について対比して表わす。 また、 この両釉に ついての色差 Δ Ε * は 0 . 4 7であり、 ステップ S 8 0で合格判定されたこと が判る。  FIG. 29 shows the result of the computer color matching method according to the fourth embodiment. The glaze prepared at the final blending ratio determined for each pigment and the target color sample glaze are shown in FIG. The stimulus values are shown in comparison. The color difference ΔΕ * for both glazes was 0.47, indicating that the pass was judged in step S80.
以上の第 4実施例に関しては, 次のような変形も可能である。  Regarding the fourth embodiment described above, the following modifications are also possible.
例えば、 数式 5 0の費用関数 Fに替わって、 以下の数式 5 1で表わされる费用 関数 Fを用いてもよい。
Figure imgf000055_0001
For example, instead of the cost function F of Equation 50, a utility function F represented by the following Equation 51 may be used.
Figure imgf000055_0001
更には、 ステップ S 7 6にて顔料をそれぞれ微量ずつ追加増量した微係数を計 算する際に、 各顔料の追加増量調合量の最低値を最低追加増量調合量△ C step- として予め規定しておき、 この△ C stepの整数倍ずつ各顔料を追加増量するよ う構成することもできる。 そして、 このように各顔料を追加増量していき、 最も 目標色見本釉との色差 Δ Ε * が小さくなるときの追加増量量を、 微係数算出の 際の補正量とし、 この補正量を加味した調合割合を調合済みの釉を再調合する際 の最終的な調合割合とすればよい。 なお、 この場合の最低追加増量調合量 A C stepは、 各顔料を追加したときにその呈する色が僅かに変わる最小単位の追加 増量量として規定される。 また、 この場合には、 各顔料の追加増量量の最大許容 量 A Craax を予め定めておき、 各顔料の追加増量の総量をこの A Cmax で規定 すればよい。 そして、 この最大許容量 A Cmax は、 その顔料を追加したときに その呈する色が大きく変わり、 他の顔料の追加堉量では元の色への戻りができな いと思われる追加増量として規定すればよレ、。 In addition, when calculating the differential coefficient of each additional pigment by a small amount in step S76, the minimum value of the additional additional compounding amount of each pigment is defined in advance as the minimum additional additional compounding amount △ C step-. In addition, it is also possible to configure so that each pigment is additionally increased by an integral multiple of this △ C step. Each pigment is additionally added in this way, and the additional amount when the color difference Δ Ε * from the target color sample glaze becomes the smallest is used as a correction amount for calculating the differential coefficient, and this correction amount is taken into account. The blending ratio determined may be used as the final blending ratio when re-mixing the blended glaze. In this case, the minimum additional compounding amount AC The step is defined as the minimum increment of the minimum unit in which the color changes slightly when each pigment is added. In this case, the maximum allowable amount A Craax of the additional amount of each pigment may be determined in advance, and the total amount of the additional amount of each pigment may be defined by this A Cmax. The maximum allowable amount A Cmax is defined as an additional increase in the color that the color changes when the pigment is added, and it is considered that the original color cannot be returned with the additional amount of the other pigment. Yeah.
また、 上記の费用関数 Fを用いた線形計画法に限られるわけではなく、 他の手 法を採ることもできる。 例えば、 上記の第 4実施例で示したように、 色差 Δ Ε * で合否判定する場合には、 その合否判定にある程度の幅が許容される。 従って、 この合否判定にある程度の瞹眛さを導入したいわゆるファジィ線形計画法の手法 を採ることもできる。 このフアジィ«計西肤の手法を採ったコンピュータカラー マッチング方法での結果を以下の図 3 0に示す。 この図 3 0に示すように、 各顔 料について定めた最終的な調合割合で調合した釉と目標色見本釉とは、 その色差 Δ Ε * が 0. 2 0であり、 両者の色はよく一致していることが判る。  Further, the present invention is not limited to the linear programming using the above-mentioned function F, and other methods can be adopted. For example, as shown in the fourth embodiment, when the pass / fail judgment is made based on the color difference ΔΕ *, a certain width is allowed for the pass / fail judgment. Therefore, a so-called fuzzy linear programming method that introduces a certain degree of ambiguity in this pass / fail decision can be adopted. The result of the computer color matching method using the method of Fujii «Keisei 肤 is shown in Figure 30 below. As shown in Fig. 30, the glaze blended at the final blending ratio determined for each pigment and the target color sample glaze have a color difference ΔΕ * of 0.20, and both colors are well It turns out that they match.
また、 上記の第 4実施例では、 陶器やタイルを色付けする釉を例に採り説明し たが、 繊維を染める染色剤にも適用できることは勿論である。  Further, in the fourth embodiment, the glaze for coloring pottery and tiles has been described as an example, but it is needless to say that the present invention can be applied to a dye for dyeing fibers.
[産業上の利用可能性] [Industrial applicability]
この発明にかかるコンピュータカラーマツチング方法および装置は、 陶器ゃタ ィルを色付けするための釉に混合される着色剤の調合割合の予測とその混合物 A computer color matching method and apparatus according to the present invention include a method for predicting a mixing ratio of a colorant to be mixed with a glaze for coloring pottery tiles and a mixture thereof.
(すなわち釉) の色予測に適用できる他に, 繊維を染めるための染色剤等の各種 の混合物における着色剤の調合割合の予測とその混合物の色予測等に適用可能で ある。 In addition to being applicable to the color prediction of (glaze), it can be applied to the prediction of the blending ratio of the colorant in various mixtures such as dyes for dyeing fibers and the color prediction of the mixture.

Claims

請求の範囲  The scope of the claims
1 . コンピュータカラ一マッチングによって着色剤の調合割合の予測または混合 物の色予測を行なう方法であって、 1. A method of estimating the mixing ratio of a colorant or the color of a mixture by computer color matching,
無色透明でない被着色物に白色着色剤を混合した第 1の混合物の散乱係数 s w. , を、 前記白色着色剤の調合率 cw に依存した関数 f ( Cw ) として準備するェ 程と、 Preparing a scattering coefficient s w ., Of a first mixture obtained by mixing a white colorant with a colorless and transparent material as a function f (C w ) depending on the mixing ratio c w of the white colorant; ,
前記散乱係数 s w, を基準として、 前記第 1の混合物の吸収係数 K w, を前記調 合率 c„ に依存する形式で求める工程と、 Obtaining an absorption coefficient K w , of the first mixture in a form depending on the mixing rate c „, based on the scattering coefficient s w ,
前記散乱係数 s w' を基準として、 白色でない有色着色剤の吸収係数 κ ,, と散 乱係数 s p とを前記有色着色剤の調合率 cp に依存する形式で求める工程と、 所望の色を有する混合物を調整するための着色剤の調合割合、 または、 所定の 調合割合で生成される混合物の色を、 前記吸収係数 Kw' , K p および前記散乱 係数 sw' , S P を用いたコンピュータカラーマッチングを行なうことによって 求める工程と、 The basis of the scattering coefficient s w ', a step of determining a format which depends the absorption coefficient kappa ,, and the scattering coefficient s p colored colorant not white in formulation ratio c p of the colored colorant, desired color formulation proportion of the coloring agent in order to adjust the mixture with use or the color of the mixture produced in a predetermined compounding ratio, the absorption coefficient K w ', K p and the scattering coefficient s w', the S P, The process required by performing computer color matching,
を備えるコンピュータカラ一マッチング方法。 A computer color matching method comprising:
2. 請求項 1記載の コンピュータカラーマッチング方法であって、 2. The computer color matching method according to claim 1, wherein
前記第 1の混合物の散乱係数 sw' を準備する工程は、 Preparing a scattering coefficient s w ′ of the first mixture,
( a ) 前記被着色物に前記白色着色剤を混合して、 前記白色着色剤の調合率 C w が異なる複数個の第 1の混合物を作成するとともに、 前記複数個の第 1の混合物 の分光反射率をそれぞれ測定する工程と、 (A) said mixing the white colorant to be colored object, the spectral of the with Formulation ratio C w of the white colorant to create a plurality of different first mixture, the first mixture of said plurality Measuring the reflectivity,
( b ) 前記被着色物に前記有色着色物を混合した第 2の混合物を作成するととも に、 前記第 2の混合物の分光反射率を測定する工程と、  (b) a step of preparing a second mixture in which the colored object is mixed with the colored object, and measuring a spectral reflectance of the second mixture;
( c ) 前記被着色物に前記白色着色剤と前記有色着色剤とを混合して、 前記有色 着色剤の調合率 Cp が異なる複数個の第 3の混合物を作成するとともに、 前記複 数個の第 3の混合物の分光反射率をそれぞれ測定する工程と、 (d) 前記被着色物に前記白色着色剤と前記有色着色剤とを混合して、 前記第 3 の混合物とは調合割合が異なる第 4の混合物を作成するとともに、 前記第 4の混 合物の分光反射率を測定する工程と、 (C) said mixing with the white colorant to be colored object and the colored colorant, along with Formulation ratio C p of the colored colorant to create a plurality of different third mixture, several said double Measuring the spectral reflectances of the third mixture of (d) mixing the white colorant and the colored colorant with the object to be colored to form a fourth mixture having a different mixing ratio from the third mixture, and the fourth mixture. Measuring the spectral reflectance of the
(e) 前記複数個の第 1の混合物の散乱係数 Sw' を前記白色着色剤の調合率 Cw の関数 f (Cw ) によって表わすとともに、 前記関数 f (Cw ) に含まれる係数 の値を仮決定する工程と、 (e) The scattering coefficient S w ′ of the plurality of first mixtures is represented by a function f (C w ) of the mixing ratio C w of the white colorant, and a coefficient f included in the function f (C w ) Tentatively determining a value;
(f ) 前記複数個の第 1の混合物の分光反射率の測定値と前記関数 f (Cw ) と を用いて、 前記調合率 cw に依存した形式で前記第 1の混合物の吸収係数 Kw' を求める工程と、 (f) Using the measured values of the spectral reflectances of the plurality of first mixtures and the function f (C w ), the absorption coefficient K of the first mixture in a form dependent on the blending rate c w w '
(g) 前記第 2の混合物の分光反射率と、 前記複数個の第 3の混合物の分光反射 率と、 前記関数 f (Cw ) と、 前記吸収係数 Kw' とを用いて、 前記調合率 Cp に依存した形式で前記有色着色剤の吸収係数 Kp と散乱係数 S Ρ とをそれぞれ 求める工程と、 (g) using the spectral reflectance of the second mixture, the spectral reflectances of the plurality of third mixtures, the function f (C w ), and the absorption coefficient K w ′, Determining the absorption coefficient K p and the scattering coefficient S の of the colored colorant in a form dependent on the rate C p ;
(h) 前記吸収係数 Kw, , ΚΡ および前記散乱係数 Sw, , Sp を用いて、 前記 第 4の混合物に関するコンピュータカラーマッチングを行なうとともに、 前記コ ンピュータカラーマッチングによる予測精度を向上させるように前記関数 f (C B ) に舍まれる係数を修正する工程と、 (h) the absorption coefficient K w,, kappa [rho and the scattering coefficient S w,, using S p, performs a computer color matching for said fourth mixture, to improve the prediction accuracy by the computer color matching Correcting the coefficient included in the function f (C B ),
( i ) 前記工程 (f ) ないし (h) を繰り返すことによって前記関数 f (Cw ) に舍まれる係数を決定する工程と、 (i) determining a coefficient included in the function f (C w ) by repeating the steps (f) to (h);
を備えるコンピュータカラーマッチング方法。 Computer color matching method comprising:
3. 前記関数 f (CH ) が定数 ai と係数 SB とを舍む次の式で与えられる、 請 求項 1または 2記載のコンピュータカラーマッチング方法: 3. The computer color matching method of claim 1 or 2, wherein the function f (C H ) is given by the following equation that includes a constant ai and a coefficient S B :
f(Cw ) = (CH +a, -SB )/(Cw +a, ). f (C w ) = (C H + a, -S B ) / (C w + a,).
4. 請求項 1記載のコンピュータカラ一マッチング方法であって、 前記コン ピュータカラーマッチングを行なう T-程は, (a) 複数の着色剤を混合して、 調合率が互いに異なる複数のサンプルを準備す る工程と、 4. The computer color matching method according to claim 1, wherein the T-step of performing the computer color matching comprises: (a) mixing a plurality of coloring agents to prepare a plurality of samples having different mixing ratios;
(b) 前記複数のサンプルの分光反射率をそれぞれ測定するとともに、 前記分光 反射率の測定値から、 前記複数のサンプルのそれぞれの色を表わす所定の表色系 の座標値の実測値を求める工程と、  (b) measuring the spectral reflectances of the plurality of samples, respectively, and obtaining actual measured values of coordinate values of a predetermined color system representing respective colors of the plurality of samples from the measured values of the spectral reflectances; When,
(c) 前記複数のサンプルのそれぞれに関して、 前記吸収係数 Kw' , KP およ び前記散乱係数 Sw' , Sp を用いて前記表色系の座標値の予測値を求めるとと もに, 前記表色系の座標値の予測誤差を算出する工程と、 (c) for each of said plurality of samples, the absorption coefficient K w ', K P and the scattering coefficient S w', also when using the S p obtain the prediction value of the coordinate values of the color system Calculating a prediction error of the coordinate values of the color system;
( d ) 前記複数のサンプルに関する前記表色系の座標储と前記予測誤差との関係 を、 所定の誤差補正法で分析する工程と、  (d) analyzing the relationship between the coordinate 储 of the color system and the prediction error for the plurality of samples by a predetermined error correction method,
(e) 前記誤差補正法を用いてコンピュータカラーマッチングの目標値または予 測値を補正しつつ、 新たな混合物の着色剤の調合割合の予測または混合物の色予 測を, 前記吸収係数 Kw' , P および前記散乱係数 Sw' , Sp を用いて求める 工程と、 (e) While correcting the target value or the predicted value of the computer color matching using the error correction method, the prediction of the mixing ratio of the colorant of the new mixture or the color prediction of the mixture is performed by the absorption coefficient K w ′. , P and the scattering coefficients S w ′, S p ,
を備えるコンピュータカラ一マッチング方法。 A computer color matching method comprising:
5. 請求項 1記載のコンピュータカラーマッチング方法であって、 前記コン ピュータカラーマツチングを行なう工程は, 5. The computer color matching method according to claim 1, wherein the step of performing the computer color matching comprises:
( a ) 着色剤の調合率が既知で前記所望の色に近い色を有する近接色サンプルに ついて所定の表色系の座標の実測値を求める工程と、  (a) obtaining an actual measurement value of coordinates of a predetermined color system for an adjacent color sample having a colorant formulation known and having a color close to the desired color;
(b) 前記近接色サンプルの既知の調合割合から、 前記吸収係数 Kw' , Kp お よび前記散乱係数 Sw' , Sp を用いて前記近接色サンプルの色を表わす前記表 色系の座標の計算値を求め、 前記実測値と前記計算値から計算誤差を求める工程 と、 (b) from a known compounding ratio of the proximity color sample, the absorption coefficient K w ', K p Contact and the scattering coefficient S w', of the color system representing the color of the adjacent color sample using the S p Obtaining a calculated value of the coordinates, obtaining a calculation error from the measured value and the calculated value;
(c) 目標混合物の色に対する前記表色系の座標の目標値を設定する工程と、 (c) setting a target value of the coordinates of the color system for the color of the target mixture;
(d) 前記計算誤差を用いて前記目標値を補正し、 補正後の目標値と前記吸収係 数 Κ„' , ΚΡおよび前記散乱係数 S„' , Spとを用いたコンピュータカラーマツ チングによって、 前記目標混合物の着色剤の調合割合を予測する工程と、 を備えるコンピュータカラーマッチング方法。 ; (d) using the calculated error by correcting the target value, the corrected target value and the absorption coefficient number kappa of "', kappa [rho and the scattering coefficient S"', computer color pine using the S p Estimating the blending ratio of the colorant in the target mixture by ching.
6 . 請求項 1記載のコンピュータカラーマッチング方法であって, 前記コン ピュータカラーマツチングを行なう工程は, 6. The computer color matching method according to claim 1, wherein the step of performing the computer color matching comprises:
複数の着色剤を調合した調合物が所望の目標色に近似した色を呈するように該 複数の着色剤についての調合割合を求める工程を舍み,  Providing a process of determining a blending ratio of the plurality of colorants so that the blended product of the plurality of colorants exhibits a color similar to a desired target color;
前記調合割合を求める工程は,  The step of obtaining the blending ratio comprises:
( a ) 前記目標色を呈する調合物見本について、 所定の表色系での色評価値の実 測値を求める丁-程と、  (a) obtaining a measured value of a color evaluation value in a predetermined color system for a preparation sample exhibiting the target color;
( b ) 既知の調合割合で前記着色剤が調合された 1次調合物について、 前記所定 の表色系での色評価値の実測値を求める工程と、  (b) a step of obtaining an actual measurement value of a color evaluation value in the predetermined color system, for a primary preparation in which the colorant is prepared at a known preparation ratio,
( c ) 前記〗次調合物についての前記既知の調合割合に基づいて、 前記 1次調合 物の呈する色の前記所定の表色系での色評価値の計算値を, 前記吸収係数 Kw' , K p および前記散乱係数 S w' , S p を用いて求める工程と、 (d ) 前記 1次調 合物に前記着色剤を増量補正したと仮定した着色剤増量調合物についての前記所 定の表色系での色評価値の計算値を, 前記吸収係数 Kw' , K p および前記散乱 係数 S w' , S p を用いて求めるとともに、 前記 1次調合物から前記着色剤増量 調合物への前記色評価値の計算値の変化量を求める工程と、 (c) calculating the calculated color evaluation value of the color of the primary formulation in the predetermined color system based on the known formulation ratio of the〗 -order formulation, using the absorption coefficient K w ′ , K p and the scattering coefficients S w ′, S p , and (d) the above-mentioned specifications for the colorant-extended formulation assuming that the colorant was increased in the primary formulation. of the calculated values of the color evaluation value in a color system, the absorption coefficient K w ', K p and the scattering coefficient S w', with obtained using S p, the colorant bulking formulation from the primary formulation Determining the amount of change in the calculated value of the color evaluation value for the object;
( e ) 前記調合物見木と前記 1次調合物との前記色評価値の実測値の差が所定範 囲で一致するように、 前記色評価値の計算値の変化量に基づいて前記着色剤のそ れそれの増量補正量を算出する工程と、  (e) The coloring based on the change in the calculated value of the color evaluation value so that the difference between the measured value of the color evaluation value of the mixture look-ahead and the primary formulation is equal within a predetermined range. Calculating the respective amount-increasing correction amounts of the agent;
を備えるコンピュータカラーマッチング方法。 Computer color matching method comprising:
7 . コンピュータカラーマッチングによつて着色剤の調合割合の予測または混合 物の色予測を行なう方法であって、 7. A method for predicting the mixing ratio of a colorant or the color of a mixture by computer color matching,
( a ) 複数の着色剤を混合して、 調合率が互いに異なる複数のサンプルを準備す る工程と、 (a) Prepare multiple samples with different blending ratios by mixing multiple colorants Process,
( b ) 前記複数のサンプルの分光反射率をそれぞれ測定するとともに、 前記分光 反射率の測定値から、 前記複数のサンプルのそれぞれの色を表わす所定の表色系 の座標値の実測値を求める工程と、  (b) a step of measuring the spectral reflectances of the plurality of samples, respectively, and obtaining an actual measurement value of a coordinate value of a predetermined color system representing each color of the plurality of samples from the measured values of the spectral reflectances; When,
( c ) 前記複数のサンプルのそれぞれに関して、 前記表色系の座標値の予測誤差 を算出するて程と、  (c) calculating a prediction error of coordinate values of the color system for each of the plurality of samples;
( d ) 前記複数のサンプルに関する前記表色系の座標値と前記予測誤差との関係 を、 所定の誤差補正法で分析する工程と、  (d) analyzing the relationship between the coordinate value of the color system and the prediction error for the plurality of samples by a predetermined error correction method,
( e ) 前記誤差補正法を用いてコンピュータカラーマッチングの目標値または予 測値を補正しつつ、 新たな混合物の着色剤の調合割合の予測または混合物の色予 測をコンピュータカラーマッチングにより行なう T-程と、  (e) While correcting the target value or the predicted value of the computer color matching using the error correction method, the prediction of the mixing ratio of the colorant in the new mixture or the color prediction of the mixture is performed by the computer color matching. About
を備えるコンピュータカラ一マッチング方法。 A computer color matching method comprising:
8 . 請求項 7記載のコンピュータカラーマッチング方法であって、 8. The computer color matching method according to claim 7, wherein
前記工程 (d ) は、 前記複数のサンプルに関する前記表色系の座標値と前記予 測誤差との関係をニューラルネッ卜ワークに学習させる工程、 を舍み、  The step (d) comprises a step of causing a neural network to learn a relationship between the coordinate values of the color system and the prediction error for the plurality of samples.
前記工程 (e ) は、 学習済みのニューラルネットワークを用いてコンピュータ カラ一マツチングによる予測を行なう工程を舍む、  The step (e) includes a step of performing prediction by computer matching using a trained neural network.
コンピュータカラーマッチング方法。 Computer color matching method.
9 . 請求項 8記載のコンピュータカラーマツチング方法であって、 9. The computer color matching method according to claim 8, wherein
前記ニューラルネットワークは、 3つのニューロンで構成される入力層と、 複 数のニューロンを舍む中間層と、 3つのニューロンで構成される出力層と、 で構 成される三層の階層構造を有する、 コンピュータカラーマッチング方法。  The neural network has a three-layer hierarchical structure composed of an input layer composed of three neurons, an intermediate layer composed of a plurality of neurons, and an output layer composed of three neurons. , Computer color matching method.
1 0 . 所望の色を有する目標混合物の着色剤の調合割合をコンピュータカラ一マツ チングによって予測する方法であって、 ( ) 着色剤の調合率が既知で前記所望の色に近い色を有する近接色サンプルに ついて所定の表色系の座標の実測値を求める工程と、 10. A method for predicting the blending ratio of a colorant in a target mixture having a desired color by computer color matching, (D) obtaining the actual measurement value of the coordinates of a predetermined color system for a nearby color sample having a colorant blending ratio known and having a color close to the desired color;
( b ) 前記近接色サンプルの既知の調合割合から、 前記近接色サンプルの色を表 わす前記表色系の座標の計算値を求め、 前記実測値と前記計算値から計算誤差を 求める工程と、  (b) obtaining a calculated value of the coordinates of the color system representing the color of the adjacent color sample from a known blending ratio of the adjacent color sample, and obtaining a calculation error from the measured value and the calculated value;
( c ) ifS目 合物の色に文 る IWti表色系の座標の目標値を設定する工程と、 ( d ) 前記計算誤差を用いて前記目標値を補正し、 補正後の目標値を用いてコン ピュータカラーマツチングを実行することによって、 前記目標混合物の着色剤の 調合割合を予測する工程と、  (c) setting a target value of the coordinates of the IWti color system written in the color of the ifS compound; and (d) correcting the target value using the calculation error, and using the corrected target value. Estimating the mixing ratio of the colorant of the target mixture by performing computer color matching with
を備えるコンピュータカラーマッチング方法。 Computer color matching method comprising:
1 1 . 請求項 1 0記載のコンピュータカラ一マッチング方法であって、 11. The computer color matching method according to claim 10, wherein
前記工程 (a ) は、 複数のサンプルに関して、 着色剤の調合割合と、 前記表色 系の座標の実測値とを舍むデータベースから、 前記目標混合物との色差が最小と なるサンプル ¾1択することによって前記近接色サンプルを検索する工程を含む、 コンピュータカラーマッチング方法。  In the step (a), for a plurality of samples, a sample that minimizes the color difference from the target mixture is selected from a database that stores the mixing ratio of the colorant and the actually measured values of the coordinates of the color system. Computer color matching method, comprising: searching for the close color sample according to:
1 2 . 複数の着色剤を調合した調合物が所望の目標色に近似した色を呈するよう に該複数の着色剤についての調合割合を求めるコンピュータカラ一マッチング方 法であって、 12. A computer color matching method for determining a blending ratio of a plurality of colorants so that a blended product of a plurality of colorants exhibits a color similar to a desired target color,
( a ) 前記目標色を呈する調合物見本について、 所定の表色系での色評価値の実 測値を求める工程と、  (a) obtaining a measured value of a color evaluation value in a predetermined color system for a preparation sample exhibiting the target color;
( b ) 既知の調合割合で前記着色剤が調合された 1次調合物について、 前記所定 の表色系での色評価値の実測値を求める工程と、  (b) a step of obtaining an actual measurement value of a color evaluation value in the predetermined color system, for a primary preparation in which the colorant is prepared at a known preparation ratio,
( c ) 前記 1次調合物についての前記既知の調合割合に基づいて、 前記 1次調合 物の呈する色の前記所定の表色系での色評価値の計算値を求める工程と、 ( d ) 前記 1次調合物に前記着色剤を増量補正したと仮定した着色剤増量調合物につい ての前記所定の表色系での色評価値の計算値を求め、 前記 1次調合物から前記着 色剤増量調合物への前記色評価値の計算値の変化量を求める工程と、 (c) a step of obtaining a calculated value of a color evaluation value of the color of the primary formulation in the predetermined color system based on the known blending ratio of the primary formulation; (d) Calculate the color evaluation value in the predetermined color system for the colorant-extended formulation assuming that the colorant has been increased in the primary formulation and calculate the color evaluation value from the primary formulation. Determining the amount of change in the calculated value of the color evaluation value to the colorant increase formulation,
( e ) 前記調合物見木と前記 1次調合物との前記色評価値の実測値の差が所定範 囲で一致するように、 前記色評価値の計算値の変化量に基づいて前記着色剤のそ れぞれの增量補正量を算出する工程と、 を備える  (e) The coloring based on the change in the calculated value of the color evaluation value so that the difference between the measured value of the color evaluation value of the mixture look-ahead and the primary formulation is equal within a predetermined range. Calculating a correction amount for each of the agents.
コンピュータカラーマッチング方法。  Computer color matching method.
1 3 . 請求項 1 2記載のコンピュータカラ一マッチング方法であって、 13. The computer color matching method according to claim 12, wherein
前記工程 (d ) は、 前記 1次調合物における前記着色剤の調合量に比べて微量 の量の前記着色剤を前記 1次調合物に増量補正したと仮定した場合について、 前 記色評価値の計算値の変化量を求める工程を舍む。  In the step (d), the color evaluation value is based on a case where it is assumed that a small amount of the coloring agent is corrected to be added to the primary formulation compared to the blending amount of the coloring agent in the primary formulation. A process for obtaining the amount of change in the calculated value of is established.
1 4 . 請求項 1 2又は請求項 1 3記載のコンピュータカラーマッチング方法であつ て、 14. The computer color matching method according to claim 12 or claim 13, wherein:
前記工程 (e ) は、 前記着色剤の増量に伴う派生費用を表わす費用関数を用い た線形計画法にて、 前記着色剤のそれぞれについての最小の増量補正量を算出す る工程を舍む。  The step (e) includes a step of calculating the minimum amount of increase correction for each of the colorants by a linear programming method using a cost function representing a derivation cost accompanying the increase in the amount of the colorant.
1 5 . コンピュータカラーマッチングによって着色剤の調合割合の予測または混 合物の色予測を行なう装置であって、 15. An apparatus for predicting the mixing ratio of colorants or predicting the color of a mixture by computer color matching,
無色透明でない被着色物に白色着色剤を混合した第】の混合物の散乱係数 S w_ ' を、 前記白色着色剤の調合率 Cw に依存した関数 f ( Cw ) として作成する手 段と、 Means for preparing the scattering coefficient S w _ 'of the second mixture obtained by mixing a white colorant with a colorless and non-transparent substance as a function f (C w ) depending on the blending ratio C w of the white colorant; ,
前記散乱係数 S を基準として、 前記第 1の混合物の吸収係数 K w, を前記調 合率 cw に依存する形式で求める手段と、 Means for determining an absorption coefficient K w , of the first mixture in a form depending on the mixing rate c w , based on the scattering coefficient S,
前記散乱係数 S を基 として、 白色でない有色着色剤の吸収係数 K P と散 乱係数 S p とを前記有色着色剤の調合率 Cp に依存する形式で求める手段と、 所望の色を有する混合物を調整するための着色剤の調合割合、 または、 所定の 調合割合で生成される混合物の色を、 前記吸収係数 Kw' , K p および前記散乱 係数 S w' , S P を用いたコンピュータカラーマッチングを行なうことによって 求める手段と、 Means for determining, on the basis of the scattering coefficient S, an absorption coefficient K P and a scattering coefficient S p of a non-white colored colorant in a form depending on the mixing rate C p of the colored colorant; and a mixture having a desired color. The proportion of the colorant to adjust the color, or The color of the mixture produced in formulation ratio, the absorption coefficient K w ', K p and the scattering coefficient S w', and means for obtaining by performing a computer color matching using S P,
を備えるコンピュータカラ一マッチング装置。 A computer color matching device comprising:
1 6 . コンピュータカラ一マッチングによって着色剤の調合割合の予測または混 合物の色予測を行なう装置であって、 16. An apparatus for predicting the mixing ratio of colorants or predicting the color of a mixture by computer color matching,
複数の着色剤を混合することによって作成された調合率が互いに異なる複数の サンプルの分光反射率をそれぞれ測定する手段と,  Means for measuring the spectral reflectances of a plurality of samples having different mixing ratios formed by mixing a plurality of colorants, respectively;
前記分光反射率の測定値から、 前記複数のサンプルのそれぞれの色を表わす所 定の表色系の座標値の実測値を求める手段と、  Means for obtaining, from the measured values of the spectral reflectance, actual measured values of coordinate values of a predetermined color system representing respective colors of the plurality of samples;
前記複数のサンプルのそれぞれに関して、 前記表色系の座標値の予測誤差を算 出する手段と、  Means for calculating a prediction error of the coordinate value of the color system for each of the plurality of samples;
前記複数のサンプルに関する前記表色系の座標値と前記予測誤差との関係を、 所定の誤差補正法で分析する丰段と、  Analyzing the relationship between the coordinate values of the color system for the plurality of samples and the prediction error by a predetermined error correction method,
前記誤差補正法を用いてコンピュータカラ一マッチングの目標値または予測値 を補正しつつ、 新たな混合物の着色剤の調合割合の予測または混合物の色予測を コンピュータカラーマッチングにより行なう手段と、  Means for correcting the target value or the predicted value of the computer color matching using the error correction method, and performing the prediction of the mixing ratio of the colorant of the new mixture or the color prediction of the mixture by computer color matching,
を備えるコンピュータカラ一マッチング装置。 A computer color matching device comprising:
1 7 . 所望の色を有する目標混合物の着色剤の調合割合をコンピュータカラーマツ チングによって予測する装置であって、 17. An apparatus for predicting, by computer color matching, the blending ratio of a colorant of a target mixture having a desired color,
着色剤の調合率が既知で前記所望の色に近い色を有する近接色サンプルについ て所定の表色系の座標の実測値を求める手段と、  Means for determining an actual measurement value of coordinates of a predetermined color system for a proximity color sample having a colorant mixing ratio known and having a color close to the desired color;
前記近接色サンプルの既知の調合割合から、 前記近接色サンプルの色を表わす 前記表色系の座標の計算値を求め、 前記実測値と前記計算値から計算誤差を求め る手段と、 前記目標混合物の色に対する前記表色系の座標の目標値を設定する手段と、 前記計算誤差を用いて前記 標値を補正し、 補正後の目標値を用いてコン ピュータカラーマッチングを実行することによって、 前記目標混合物の着色剤の 調合割合を予測する手段と、 Means for calculating a calculated value of the coordinates of the color system representing the color of the close color sample from the known blending ratio of the close color sample, and calculating a calculation error from the measured value and the calculated value; Means for setting a target value of the coordinates of the color system for the color of the target mixture; correcting the target value using the calculation error; and performing computer color matching using the corrected target value. Means for predicting the blending ratio of the colorant in the target mixture,
を備えるコンピュータカラーマッチング装置。 A computer color matching device comprising:
1 8 . 複数の着色剤を調合した調合物が所望の目檫色に近似した色を呈するよう に該複数の着色剤についての調合割合を求めるコンピュータカラーマッチング装 置であって、 18. A computer color matching apparatus for determining a blending ratio of a plurality of colorants such that a blended product of a plurality of colorants exhibits a color similar to a desired target color,
前記目標色を呈する調合物見本について、 所定の表色系での色評価値の実測値 を求める手段と、  Means for obtaining an actual measurement value of a color evaluation value in a predetermined color system for a preparation sample exhibiting the target color;
既知の調合割合で前記着色剤が調合された〗次調合物について、 前記所定の表 色系での色評価値の実測値を求める手段と、  Means for obtaining an actual measurement value of a color evaluation value in the predetermined color system, for a secondary formulation in which the colorant is formulated at a known formulation ratio,
前記〗次調合物についての前記既知の調合割合に基づいて、 前記 1次調合物の 呈する色の前記所定の表色系での色評価値の計算値を求める手段と、  Means for obtaining a calculated value of a color evaluation value of the color represented by the primary formulation in the predetermined color system, based on the known formulation ratio of the primary formulation;
前記 1次調合物に前記着色剤を増量補正したと仮定した着色剤増量調合物につ いての前記所定の表色系での色評価値の計算値を求め、 前記 1次調合物から前記 着色剤増量調合物への前記色評価値の計算値の変化量を求める手段と、  Calculating the color evaluation value in the predetermined color system for the colorant-enhanced formulation assuming that the colorant has been increased in the primary formulation and calculating the coloration from the primary formulation Means for determining the amount of change in the calculated value of the color evaluation value to the agent extension formulation,
前記調合物見本と前記 1次調合物との前記色評価値の実測値の差が所定範囲で 一致するように、 前記色評価値の計算値の変化量に基づいて前記着色剤のそれぞ れの增量補正量を算出する手段と、 を備える  Each of the coloring agents based on the amount of change in the calculated value of the color evaluation value so that the difference between the measured value of the color evaluation value of the preparation sample and that of the primary preparation matches within a predetermined range. Means for calculating the mass correction amount of
コンピュータカラ一マッチング装置。  Computer color matching device.
1 . コンピュータカラーマッチングに使用される着色剤の吸収係数と散乱係数 を決定する方法であって、 1. A method for determining the absorption and scattering coefficients of colorants used in computer color matching,
無色透明でない被着色物に白色着色剤を混合した第 1の混合物の散乱係数 S w_ ' と吸収係数 Kw' を決定する工程と、 前記散乱係数 Sw, と吸収係数 Kw, の少なくとも一方を基準として、 A色でな い有色着色剤の吸収係数 Κμ と散乱係数 S の相対値をそれぞれ決定する I:程 と、 A step of determining a scattering coefficient S w _ 'and an absorption coefficient K w ' of a first mixture obtained by mixing a white colorant with a colorless and transparent object; Based on at least one of the scattering coefficient S w , and the absorption coefficient K w , the relative values of the absorption coefficient Κ μ and the scattering coefficient S of the colorant other than A are determined.
を備える方法。 A method comprising:
20. 前記第〗の混合物の散乱係数 Sw' 力, 前記白色着色剤の調合率 ^.と定数 a, と係数 SB とを舍む次の関係で与えられる、 請求項 1 9記載の方法: 20. scattering coefficient S w 'force of the mixture of the first〗, the preparation of the white colorant ^. And constants a, and given a coefficient S B in Complex free following relationship The method of claim 1 9, wherein :
Sw, =(CW +a, -SB )/(Cw +a】 ). S w , = (C W + a, -S B ) / (C w + a]).
2 1. 前記第 1の混合物の散乱係数 Sw' 力 前記白色着色剤の調合率 Cwと定数 a, b, d, e ,CW。を含む次の関係で与えられる、 請求項 1 9記載の方法: sw, = a'Cw + b (Cwo ≤ Cw); 2 1. Scattering coefficient S w 'force of the first mixture The mixing ratio C w of the white colorant and constants a, b, d, e, and C W. 10. The method of claim 19, given by the following relation: s w , = a'C w + b (C wo ≤ C w );
Sw, = d-Cw + e (Cw < Cwo). S w , = dC w + e (C w <C wo ).
22. コンピュータカラーマッチングを利用して作成された混合釉であって、 無色透明でないベース釉に白色着色剤を混合した第 1の混合釉の散乱係数 Sw_22. Scattering coefficient S w _ of the first mixed glaze, which is a mixed glaze made using computer color matching and is a colorless and transparent base glaze mixed with a white colorant
' と吸収係数 Kw' を決定し、 'And the absorption coefficient K w '
前記散乱係数 Sw' と吸収係数 Kw' の少なくとも一方を基準として、 白色でな い有色着色剤の吸収係数 KP と散乱係数 SP の相対値をそれぞれ決定し、 前記吸収係数 Kw, , Kp および前記散乱係数 Sw, , Sp を用いたコンビュ一 タカラーマッチングを行なうことによって, 所望の色を有する混合釉を調整する ための着色剤の調合割合、 または、 所定の調合割合で生成される混合釉の色を予 測し、 Based on at least one of the scattering coefficient S w ′ and the absorption coefficient K w ′, a relative value of the absorption coefficient K P and the scattering coefficient S P of the non-white colorant is determined, and the absorption coefficient K w , , K p and the scattering coefficient S w, by performing Konbyu one Takara over matching using S p, formulation ratio of the colorant to adjust the mixed glaze having a desired color or a predetermined compounding ratio Predict the color of the mixed glaze generated by
前記予測された調合割合または前記予測された色を有するように前記ベース釉 と前記白色着色剤と前記有色着色剤とを混合する,  Mixing the base glaze, the white colorant, and the colored colorant to have the predicted blending ratio or the predicted color;
ことによって生成された混合釉。 Mixed glaze produced by.
2 3 . コンピュータカラーマッチングを利用して作成された混合釉を用いて製造 された陶磁器であって、 2 3. Ceramics manufactured using mixed glaze made using computer color matching.
無色透明でないベース釉に白色着色剤を混合した第 1の混合釉の散乱係数 S w_ ' と吸収係数 Kw, を決定し、 The scattering coefficient S w _ 'and the absorption coefficient K w , of the first mixed glaze in which the white colorant is mixed with the colorless and transparent base glaze are determined,
前記散乱係数 S w' と吸収係数 K w' の少なくとも一方を基準として、 白色でな い有色着色剤の吸収係数 K P と散乱係数 S P の相対値をそれぞれ決定し、 前記吸収係数 K , K p および前記散乱係数 S w, , を用いたコンビユー タカラーマッチングを行なうことによって, 所望の色を有する混台釉を調整する ための着色剤の調合割合、 または、 所定の調合割合で生成される混合釉の色をザ' 測し、 Based on at least one of the scattering coefficient S w ′ and the absorption coefficient K w ′, the relative values of the absorption coefficient K P and the scattering coefficient S P of the colorant that is not white are determined, and the absorption coefficients K and K are determined. By performing a combi- ter color matching using p and the scattering coefficients S w ,, a mixture ratio of a colorant for adjusting a mixed-plate glaze having a desired color or a predetermined mixture ratio is generated. Measure the color of the mixed glaze,
前記予測された調合割合または前記予測された色を有するように前記ベース釉 と前記白色着色剤と前記有色着色剤とを混合する, ことによって生成された混合 釉で陶磁器の素地の少なくとも一部が覆われた, 陶磁器。  Mixing the base glaze, the white colorant and the colored colorant so as to have the predicted blending ratio or the predicted color, wherein at least a portion of the ceramic body is Covered ceramic.
2 4 . コンピュータカラーマッチングに使用される着色剤の吸収係数と散乱係数 を決定する方法であって、 24. A method for determining the absorption and scattering coefficients of colorants used in computer color matching,
無色透明でない被着色物に基準となる顔料を混合した第 1の混合物の散乱係数 s と吸収係数 Kw' を決定する工程と、 A step of determining a scattering coefficient s and an absorption coefficient K w ′ of a first mixture obtained by mixing a reference pigment with a colorless and non-transparent object
前記散乱係数 S と吸収係数 K の少なくとも一方を基準として、 有色着色 剤の吸収係数 K P と散乱係数 S P の相対値をそれぞれ決定する工程と、 を備える方法。 The method comprising, based on the at least one of the scattering coefficient S and the absorption coefficient K, and determining the relative value of the absorption coefficient K P and the scattering coefficient S P output colored colorant, respectively, the.
2 5 . コンピュータカラーマッチングに使用される着色剤の吸収係数と散乱係数 を決定する装置であって、 25. A device for determining the absorption coefficient and scattering coefficient of a colorant used in computer color matching,
無色透明でない被着色物に基 となる顔料を混合した第 1の混合物の散乱係数 S H' と吸収係数 Kw' を決定する手段と、 Means for determining a scattering coefficient S H ′ and an absorption coefficient K w ′ of a first mixture obtained by mixing a base pigment that is not colorless and transparent;
前記散乱係数 s w' と吸収係数 K w' の少なくとも一方を基準として、 有色着色 剤の吸収係数 Κ,, と散乱係数 S,> の相対値をそれぞれ決定する上程と、 を備える方法。 Colored coloring based on at least one of the scattering coefficient s w ′ and the absorption coefficient K w ′ Determining the relative values of the absorption coefficient Κ ,, and the scattering coefficient S,> of the agent, respectively.
2 6. コンピュータカラーマッチングを利用して作成された混合釉を用いて製造 された便器。 2 6. A toilet made using mixed glaze made using computer color matching.
PCT/JP1996/000738 1996-02-22 1996-03-21 Computer color matching method and apparatus WO1997031247A6 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
JP52996597A JP3870421B2 (en) 1996-02-22 1996-03-21 Computer color matching method and apparatus

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
JP8061792A JPH09229773A (en) 1996-02-22 1996-02-22 Computerized color matching method
JP8/61792 1996-02-22

Publications (2)

Publication Number Publication Date
WO1997031247A1 true WO1997031247A1 (en) 1997-08-28
WO1997031247A6 WO1997031247A6 (en) 1997-08-28

Family

ID=

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2011059059A (en) * 2009-09-14 2011-03-24 Seiren Co Ltd Computer color matching method, and computer readable recording medium recording computer color matching program
KR101993752B1 (en) * 2018-02-27 2019-06-27 연세대학교 산학협력단 Method and Apparatus for Matching Colors Using Neural Network
WO2019218549A1 (en) * 2018-05-18 2019-11-21 Zhang Ye Intelligent color identification and coloring system and method
EP3605202A1 (en) * 2018-07-31 2020-02-05 Essilor International Method and system for determining a lens of customized color
US11062479B2 (en) 2017-12-06 2021-07-13 Axalta Coating Systems Ip Co., Llc Systems and methods for matching color and appearance of target coatings

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH01129123A (en) * 1987-11-13 1989-05-22 Kanebo Ltd Determining method of dyeing formulation in computer color matching
JPH0346525A (en) * 1989-07-14 1991-02-27 Kanebo Ltd Computer color matching method
JPH0423532A (en) * 1990-05-17 1992-01-27 Nec Corp Subcarrier multiplex optical transmission system
JPH0527813B2 (en) * 1984-07-24 1993-04-22 Toyo Ink Mfg Co
JPH05125673A (en) * 1991-11-05 1993-05-21 Kanebo Ltd Method for dyeing set processed article and method for predicting change of chromaticity following set processing
JPH08105797A (en) * 1994-10-04 1996-04-23 Toto Ltd Computer color matching method

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH0527813B2 (en) * 1984-07-24 1993-04-22 Toyo Ink Mfg Co
JPH01129123A (en) * 1987-11-13 1989-05-22 Kanebo Ltd Determining method of dyeing formulation in computer color matching
JPH0346525A (en) * 1989-07-14 1991-02-27 Kanebo Ltd Computer color matching method
JPH0423532A (en) * 1990-05-17 1992-01-27 Nec Corp Subcarrier multiplex optical transmission system
JPH05125673A (en) * 1991-11-05 1993-05-21 Kanebo Ltd Method for dyeing set processed article and method for predicting change of chromaticity following set processing
JPH08105797A (en) * 1994-10-04 1996-04-23 Toto Ltd Computer color matching method

Cited By (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2011059059A (en) * 2009-09-14 2011-03-24 Seiren Co Ltd Computer color matching method, and computer readable recording medium recording computer color matching program
US11062479B2 (en) 2017-12-06 2021-07-13 Axalta Coating Systems Ip Co., Llc Systems and methods for matching color and appearance of target coatings
US11568570B2 (en) 2017-12-06 2023-01-31 Axalta Coating Systems Ip Co., Llc Systems and methods for matching color and appearance of target coatings
US11692878B2 (en) 2017-12-06 2023-07-04 Axalta Coating Systems Ip Co., Llc Matching color and appearance of target coatings based on image entropy
KR101993752B1 (en) * 2018-02-27 2019-06-27 연세대학교 산학협력단 Method and Apparatus for Matching Colors Using Neural Network
WO2019218549A1 (en) * 2018-05-18 2019-11-21 Zhang Ye Intelligent color identification and coloring system and method
EP3605202A1 (en) * 2018-07-31 2020-02-05 Essilor International Method and system for determining a lens of customized color
WO2020025595A1 (en) * 2018-07-31 2020-02-06 Essilor International Method and system for determining a lens of customized color
CN112534342A (en) * 2018-07-31 2021-03-19 依视路国际公司 Method and system for determining custom colored lenses
CN112534342B (en) * 2018-07-31 2022-11-01 依视路国际公司 Method and system for determining custom colored lenses

Also Published As

Publication number Publication date
CN1214240C (en) 2005-08-10
CN1207808A (en) 1999-02-10
JP3870421B2 (en) 2007-01-17
JPH09229773A (en) 1997-09-05

Similar Documents

Publication Publication Date Title
CA2434554C (en) Computer-implemented neural network color matching formulation system
CN101680805B (en) Method for color matching
JP6703639B1 (en) Paint manufacturing method and method for predicting color data
CN101506633B (en) Method for color matching
EP2098840B1 (en) Texture map of coated color, its making method, making program, making system and data structure
CA2434555C (en) Computer-implemented neural network color matching formulation applications
EP0637731A1 (en) Method of color reproduction, method of reproducing a coating color, and method of selecting a coating color
AU2004248204B2 (en) Recipe calculation method for matt colour shades
Dupont Study of the reconstruction of reflectance curves based on tristimulus values: comparison of methods of optimization
JP3870421B2 (en) Computer color matching method and apparatus
WO2021132654A1 (en) Method for producing paint material, method for predicting hue data, and computer color-mixing system
JPWO1997031247A6 (en) Computer color matching method and apparatus
US7167246B1 (en) Method of color matching metallic paints
JP2006030210A (en) Method and system of computer color matching
CN113362194A (en) Coating quality prediction device and method for generating learned model
Zhao et al. Predicting the spectral reflectance factor of translucent paints using Kubelka‐Munk turbid media theory: Review and evaluation
JP3440597B2 (en) Computer color matching method
JP6936416B1 (en) How to make paint and how to predict color data
JPH0894442A (en) Computerrized color matching method
JP3836120B1 (en) Computer color matching method, color matching apparatus and program
JP2003294530A (en) Calculation method of computer color matching
JPS6161042A (en) Method of characterizing optical property of surface containing metallic particle by apparatus
JPH05296836A (en) Method and apparatus for color matching
JP3509227B2 (en) Computer color matching method
CN114174784A (en) Method and system for quantifying spectral similarity between a sample color and a target color

Legal Events

Date Code Title Description
WWE Wipo information: entry into national phase

Ref document number: 96199756.7

Country of ref document: CN

AK Designated states

Kind code of ref document: A1

Designated state(s): CN JP US

DFPE Request for preliminary examination filed prior to expiration of 19th month from priority date (pct application filed before 20040101)