JP2005142621A - Image processing apparatus, method, and program - Google Patents

Image processing apparatus, method, and program Download PDF

Info

Publication number
JP2005142621A
JP2005142621A JP2003374097A JP2003374097A JP2005142621A JP 2005142621 A JP2005142621 A JP 2005142621A JP 2003374097 A JP2003374097 A JP 2003374097A JP 2003374097 A JP2003374097 A JP 2003374097A JP 2005142621 A JP2005142621 A JP 2005142621A
Authority
JP
Japan
Prior art keywords
pixel
visible
data
value
defect depth
Prior art date
Legal status (The legal status 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 status listed.)
Granted
Application number
JP2003374097A
Other languages
Japanese (ja)
Other versions
JP3997486B2 (en
Inventor
Koji Kita
耕次 北
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Noritsu Koki Co Ltd
Original Assignee
Noritsu Koki Co Ltd
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 Noritsu Koki Co Ltd filed Critical Noritsu Koki Co Ltd
Priority to JP2003374097A priority Critical patent/JP3997486B2/en
Publication of JP2005142621A publication Critical patent/JP2005142621A/en
Application granted granted Critical
Publication of JP3997486B2 publication Critical patent/JP3997486B2/en
Anticipated expiration legal-status Critical
Expired - Lifetime legal-status Critical Current

Links

Images

Landscapes

  • Facsimile Image Signal Circuits (AREA)
  • Studio Devices (AREA)
  • Image Input (AREA)
  • Image Processing (AREA)
  • Facsimiles In General (AREA)
  • Color Television Image Signal Generators (AREA)
  • Facsimile Scanning Arrangements (AREA)

Abstract

<P>PROBLEM TO BE SOLVED: To provide an image processing apparatus, method, and program for accurately correcting a deviation between the effect of a defect on visible data and the effect of a defect on invisible data and reducing a computation processing amount for the correction. <P>SOLUTION: A target pixel is selected from defective pixels, a defect depth DF<SB>m, n</SB>based on invisible data of the target pixel is calculated, average pixel values Rave<SB>m, n</SB>, Gave<SB>m, n</SB>, Bave<SB>m, n</SB>, of normal pixels around the target pixel are calculated, defect depth Rf<SB>m, n</SB>, Gf<SB>m, n</SB>, Bf<SB>m, n</SB>based on visible data of the target pixel are calculated, the defect depth of invisible and visible data is stored by calculation for a plurality of number of times by sequentially selecting the target pixel, regression analysis is applied to them to calculate regression coefficients Ra, Ga, Ba and correlation coefficients Rr, Gr, Br, a correction coefficient W to correct the deviation between the defect depth of the visible data and the defect depth of the invisible data is calculated on the basis of the regression coefficients and the correlation coefficients, the luminance of the defective pixels is adjusted to correct the defective pixel on the basis of the correction coefficient W and the invisible defect depth Df<SB>m, n</SB>of each defective pixel. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

本発明は、画像記録材料に対して可視光及び非可視光を照射して可視データ及び非可視データをそれぞれ取得し、前記非可視データに基づいて前記可視データに含まれる欠陥画素を検出するとともに、前記非可視データを用いて前記欠陥画素の輝度調整を行う画像処理装置、方法及びプログラムに関する。   The present invention obtains visible data and invisible data by irradiating the image recording material with visible light and invisible light, respectively, and detects defective pixels included in the visible data based on the invisible data. The present invention relates to an image processing apparatus, method, and program for performing luminance adjustment of the defective pixel using the invisible data.

写真フィルム等の画像記録材料に対して可視光を照射した場合、画像記録材料に記録されている画像の濃度に応じて透過光量が変化するが、傷や埃等の欠陥がある箇所でもそれらの欠陥により光が一部屈折したり反射したりすることで透過光量が変化する。これに対して、画像記録材料に対して非可視光を照射した場合には、傷や埃等の欠陥がある箇所では透過光量が変化するものの、画像記録材料に記録されている画像の濃度の影響によっては透過光量は変化しない。したがって、画像記録材料に非可視光を照射してその透過光を受光して得られる非可視データに基づいて、画像記録材料に付いている傷や埃等の欠陥を検出するとともに、この欠陥部における非可視データの変動量に基づいて、画像記録材料に記録されている画像(可視光を照射してその透過光を受光して得られる可視データ)の輝度調整を行うことにより、欠陥部の画像を修正することが可能である(例えば、特許文献1参照)。   When the image recording material such as a photographic film is irradiated with visible light, the amount of transmitted light changes depending on the density of the image recorded on the image recording material. The amount of transmitted light changes due to light being partially refracted or reflected by a defect. On the other hand, when the image recording material is irradiated with invisible light, the amount of transmitted light changes in a place where there is a defect such as a scratch or dust, but the density of the image recorded on the image recording material The amount of transmitted light does not change depending on the influence. Therefore, based on the invisible data obtained by irradiating the image recording material with invisible light and receiving the transmitted light, defects such as scratches and dust attached to the image recording material are detected. By adjusting the brightness of the image recorded on the image recording material (visible data obtained by irradiating visible light and receiving the transmitted light) based on the amount of fluctuation of invisible data in It is possible to correct the image (for example, see Patent Document 1).

ところで、画像記録材料に可視光及び非可視光を照射した場合における前記欠陥部での透過光量の変化量は、可視光と非可視光との波長或いは屈折率の相違により、可視光と非可視光とでは若干異なっている。そこで、このような可視光と非可視光との相違を補正して、輝度調整方法による欠陥部の修正精度を向上させる技術として、例えば、以下の技術が知られている。この技術は、まず、画像記録材料に可視光及び非可視光を照射したときの処理対象の欠陥部を含む一部領域のデータを各々抽出し、それらのデータをハイパスフィルタに各々入力することで可視データ及び非可視データから高周波成分データをそれぞれ生成する。次に、前記処理対象の欠陥部に属する各欠陥画素の中から処理対象の欠陥画素を選択し、前記各高周波成分データから前記選択した欠陥画素の高周波成分データを各々抽出する。そして、処理対象の欠陥画素の可視データの高周波成分データと非可視データの高周波成分データとの比(高周波成分の比)を演算し、この比を処理対象の欠陥画素の非可視データに乗じた値を、処理対象の欠陥画素における可視データの画素値の修正量とする技術が知られている(例えば、特許文献2参照)。この技術によれば、前記高周波成分の比により、可視光と非可視光との波長或いは屈折率の相違に起因する前記欠陥部での透過光量の変化量の相違を補正することができる。   By the way, when the image recording material is irradiated with visible light and non-visible light, the amount of change in the amount of transmitted light at the defect portion varies depending on the wavelength or refractive index between visible light and non-visible light. It is slightly different from light. Therefore, as a technique for correcting such a difference between visible light and invisible light and improving the correction accuracy of the defect portion by the brightness adjustment method, for example, the following techniques are known. In this technique, first, data of a partial region including a defective portion to be processed when an image recording material is irradiated with visible light and invisible light is extracted, and the data is input to a high-pass filter. High frequency component data is generated from visible data and non-visible data, respectively. Next, a defective pixel to be processed is selected from the defective pixels belonging to the defective portion to be processed, and high-frequency component data of the selected defective pixel is extracted from each high-frequency component data. Then, the ratio of the high-frequency component data of the visible data of the defective pixel to be processed and the high-frequency component data of the invisible data (ratio of high-frequency components) was calculated, and this ratio was multiplied by the non-visible data of the defective pixel to be processed. A technique is known in which a value is used as a correction amount of a pixel value of visible data in a defective pixel to be processed (see, for example, Patent Document 2). According to this technique, the difference in the amount of change in the amount of transmitted light at the defective portion due to the difference in wavelength or refractive index between visible light and invisible light can be corrected by the ratio of the high frequency components.

特開平2000−115464号公報(第5頁、第3図)Japanese Unexamined Patent Publication No. 2000-115464 (5th page, FIG. 3) 特開平2001−078038号公報(第16−17頁、第10図)Japanese Unexamined Patent Publication No. 2001-078038 (pages 16-17, FIG. 10)

しかしながら、上記特許文献2に記載の従来技術では、前記高周波成分データを生成する処理を処理対象の各欠陥部毎に行うとともに、前記高周波成分の比の演算を前記処理対象の欠陥部に属する全ての欠陥画素についてそれぞれ行うことから、演算処理量が多くなり、装置の処理時間が長くなるという問題があった。   However, in the prior art described in Patent Document 2, the processing for generating the high-frequency component data is performed for each defect portion to be processed, and the calculation of the ratio of the high-frequency components belongs to all the defect portions to be processed. Since each of the defective pixels is performed, there is a problem that the amount of calculation processing increases and the processing time of the apparatus becomes long.

また、可視データ及び非可視データからそれぞれ生成する高周波成分データは、欠陥の状態や可視データに含まれる画像の状態等による影響を大きく受けることから、各欠陥画素毎に前記高周波成分の比を演算しても、その高周波成分の比が、可視光と非可視光との波長或いは屈折率の相違に起因する前記欠陥部での透過光量の変化量の相違を正確に表すものであることは少なく、可視光と非可視光との波長或いは屈折率の相違をより正確な補正するために改善の余地があった。   In addition, the high frequency component data generated from visible data and invisible data is greatly affected by the defect state and the image state included in the visible data, so the ratio of the high frequency component is calculated for each defective pixel. Even so, it is rare that the ratio of the high-frequency components accurately represents the difference in the amount of change in the amount of transmitted light at the defect due to the difference in wavelength or refractive index between visible light and invisible light. Therefore, there is room for improvement in order to more accurately correct the difference in wavelength or refractive index between visible light and invisible light.

本発明は、上記の課題に鑑みてなされたものであり、その目的は、可視光と非可視光との波長或いは屈折率の相違に起因する、可視データが欠陥により受ける影響と非可視データが欠陥により受ける影響とのずれを正確に補正することを可能とするとともに、当該補正のための演算処理量を軽減して処理時間を短縮することが可能な画像処理装置、方法及びプログラムを提供する点にある。   The present invention has been made in view of the above-mentioned problems, and the purpose thereof is the influence of visible data on defects caused by the difference in wavelength or refractive index between visible light and invisible light, and the invisible data. Provided is an image processing apparatus, method, and program capable of accurately correcting a deviation from an effect caused by a defect and reducing the amount of calculation processing for the correction to shorten a processing time. In the point.

上記目的を達成するための本発明に係る画像処理装置の第1特徴構成は、画像記録材料に対して可視光及び非可視光を照射して可視データ及び非可視データをそれぞれ取得し、前記非可視データに基づいて検出された前記可視データに含まれる欠陥画素の修正を行う画像処理装置において、複数の前記欠陥画素の一つを対象画素として選択する対象画素選択手段と、前記非可視データの基準画素値と前記非可視データの前記対象画素に対応する画素の画素値とに基づいて前記対象画素の非可視データに基づく欠陥深度を演算する非可視欠陥深度演算手段と、前記可視データにおける前記対象画素の周囲の一定領域内の正常画素の平均画素値を演算する平均正常画素値演算手段と、前記対象画素の画素値と前記平均正常画素値とに基づいて前記対象画素の可視データに基づく欠陥深度を演算する可視欠陥深度演算手段と、前記対象画素を順次選択することにより前記非可視欠陥深度演算手段及び可視欠陥深度演算手段においてそれぞれ複数演算される前記非可視欠陥深度及び可視欠陥深度を記憶する記憶手段と、前記記憶手段に記憶された複数の前記非可視欠陥深度と前記可視欠陥深度とについて回帰分析を行い、回帰係数と相関係数とを演算する回帰演算手段と、前記回帰係数及び相関係数に基づいて前記可視データと非可視データとの欠陥深度のずれを補正する補正係数を演算する補正係数演算手段と、前記補正係数と各欠陥画素の非可視欠陥深度とに基づいて各欠陥画素の輝度調整による修正を行う修正手段とを有する点にある。
なお、前記非可視データの基準画素値は、前記画像記録材料に欠陥が存在しない理想的な状態における非可視データの画素値の平均値に基づいて演算したものを用いることができる。
In order to achieve the above object, a first characteristic configuration of an image processing apparatus according to the present invention obtains visible data and invisible data by irradiating an image recording material with visible light and invisible light, respectively. In an image processing apparatus for correcting a defective pixel included in the visible data detected based on visible data, target pixel selection means for selecting one of the plurality of defective pixels as a target pixel, and the non-visible data A non-visible defect depth calculation means for calculating a defect depth based on non-visible data of the target pixel based on a reference pixel value and a pixel value of a pixel corresponding to the target pixel of the non-visible data; Average normal pixel value calculation means for calculating an average pixel value of normal pixels in a fixed region around the target pixel, and based on the pixel value of the target pixel and the average normal pixel value Visible defect depth calculation means for calculating a defect depth based on visible data of an elephant pixel, and a plurality of non-visible defects calculated in the invisible defect depth calculation means and the visible defect depth calculation means by sequentially selecting the target pixel. A storage unit that stores a defect depth and a visible defect depth, and a regression that performs regression analysis on the plurality of invisible defect depths and the visible defect depths stored in the storage unit, and calculates a regression coefficient and a correlation coefficient Calculation means, correction coefficient calculation means for calculating a correction coefficient for correcting a defect depth shift between the visible data and non-visible data based on the regression coefficient and the correlation coefficient, non-correction of the correction coefficient and each defective pixel And correction means for correcting each defective pixel by adjusting the luminance based on the visible defect depth.
As the reference pixel value of the invisible data, a value calculated based on an average value of pixel values of the invisible data in an ideal state where the image recording material does not have a defect can be used.

この第1特徴構成によれば、複数の欠陥画素を順次対象画素として選択してそれぞれ可視欠陥深度と非可視欠陥深度とを演算し、これらの複数の可視欠陥深度と非可視欠陥深度とに基づいて回帰分析を行い、可視欠陥深度と非可視欠陥深度との関係を表す回帰係数と、この回帰係数の正確性を表す相関係数とを演算し、これらに基づいて前記可視データと非可視データとの欠陥深度のずれを補正する補正係数を演算するので、可視光と非可視光との波長或いは屈折率の相違に起因する、可視データが欠陥により受ける影響と非可視データが欠陥により受ける影響とのずれを正確に補正する輝度調整を行うことが可能となる。また、画像記録材料毎、画像毎、或いは所定の画像領域毎等の一定の範囲毎に一つの補正係数を用いて輝度調整を行うので、欠陥の補正の際の演算処理量を軽減して処理時間を短縮することが可能となる。   According to the first characteristic configuration, a plurality of defective pixels are sequentially selected as target pixels, and a visible defect depth and an invisible defect depth are calculated, respectively, and based on the plurality of visible defect depths and invisible defect depths. The regression coefficient representing the relationship between the visible defect depth and the invisible defect depth and the correlation coefficient representing the accuracy of the regression coefficient are calculated, and based on these, the visible data and the invisible data are calculated. Since the correction coefficient that corrects the deviation of the defect depth from the light is calculated, the effect of visible data on defects and the effect of non-visible data on defects due to the difference in wavelength or refractive index between visible light and invisible light Therefore, it is possible to perform brightness adjustment for accurately correcting the deviation. In addition, brightness adjustment is performed using a single correction coefficient for each image recording material, for each image, or for a predetermined range such as for each predetermined image area, thus reducing the amount of calculation processing when correcting defects. Time can be shortened.

本発明に係る画像処理装置の第2特徴構成は、上記第1特徴構成に加えて、前記補正係数演算手段が、前記回帰係数と所定の基準値との差に対して相関係数を乗算して得られた値に前記基準値を加算して補正係数を演算する点にある。   In a second feature configuration of the image processing apparatus according to the present invention, in addition to the first feature configuration, the correction coefficient calculation unit multiplies a difference between the regression coefficient and a predetermined reference value by a correlation coefficient. The correction coefficient is calculated by adding the reference value to the obtained value.

この第2特徴構成によれば、相関係数の値に応じて回帰係数が補正係数に対して与える影響を調節することになり、相関係数が高く回帰係数の正確性が高い場合には回帰係数が補正係数に与える影響を大きくし、逆に相関係数が低く回帰係数の正確性が低い場合には回帰係数が補正係数に与える影響を小さくすることができるので、前記補正係数により、前記可視データと非可視データとの欠陥深度のずれを補正する精度を高めることができる。   According to the second feature configuration, the influence of the regression coefficient on the correction coefficient is adjusted according to the value of the correlation coefficient. When the correlation coefficient is high and the accuracy of the regression coefficient is high, the regression coefficient The influence of the coefficient on the correction coefficient is increased. Conversely, when the correlation coefficient is low and the accuracy of the regression coefficient is low, the influence of the regression coefficient on the correction coefficient can be reduced. The accuracy of correcting the deviation of the defect depth between visible data and invisible data can be increased.

本発明に係る画像処理装置の第3特徴構成は、上記第1又は第2特徴構成に加えて、前記補正係数演算手段が、複数色の可視データに共通する前記補正係数を演算する場合に、各色の前記相関係数を当該色の前記回帰係数の重みとする複数色の重み付き平均値を演算して共通回帰係数とし、複数色の前記相関係数の平均値を演算して共通相関係数とし、これらの共通回帰係数及び共通相関係数を、前記回帰係数及び相関係数に代えて用いて補正係数の演算を行う点にある。   In a third feature configuration of the image processing apparatus according to the present invention, in addition to the first or second feature configuration, the correction coefficient calculation unit calculates the correction coefficient common to visible data of a plurality of colors. Calculate a weighted average value of a plurality of colors using the correlation coefficient of each color as a weight of the regression coefficient of the color to obtain a common regression coefficient, and calculate an average value of the correlation coefficients of a plurality of colors to obtain a common phase relationship The correction coefficient is calculated using the common regression coefficient and the common correlation coefficient instead of the regression coefficient and the correlation coefficient.

この第3特徴構成によれば、例えば、R(赤色)、G(緑色)、B(青色)等のように複数色の可視データが取得されている場合に、これらの複数色の可視データに共通する一つの補正係数を演算するので、可視光と非可視光との波長或いは屈折率の相違に起因する、可視データが欠陥により受ける影響と非可視データが欠陥により受ける影響とのずれを補正するための演算処理量を軽減して処理時間を短縮することができる。また、各色の前記相関係数を当該色の前記回帰係数の重みとする複数色の重み付き平均値を演算して共通回帰係数とし、複数色の可視データの中で相関係数が高く正確性が高い色の回帰係数の重みを重くすることができるので、前記補正係数を複数色の可視データについて一つしか用いないにも関わらず、前記可視データと非可視データとの欠陥深度のずれを補正する精度を高めることができる。   According to the third feature configuration, when multiple colors of visible data such as R (red), G (green), and B (blue) are acquired, the visible data of these multiple colors Since one common correction coefficient is calculated, the difference between the effect of visible data due to defects and the effect of non-visible data due to defects due to differences in wavelength or refractive index between visible light and invisible light is corrected. It is possible to reduce the amount of calculation processing for reducing the processing time. Further, a weighted average value of a plurality of colors is calculated by using the correlation coefficient of each color as the weight of the regression coefficient of the color to obtain a common regression coefficient, and the correlation coefficient is high and accurate among the visible data of the plurality of colors. Since the weight of the regression coefficient of high color can be increased, the deviation of the defect depth between the visible data and non-visible data can be reduced even though only one correction coefficient is used for visible data of a plurality of colors. The accuracy of correction can be increased.

本発明に係る画像処理装置の第4特徴構成は、上記第1から第3の何れかの特徴構成に加えて、前記平均正常画素値演算手段が、前記非可視データの基準画素値と前記非可視データの前記各正常画素に対応する画素の画素値とに基づいて前記一定領域内の各正常画素の非可視データに基づく深度修正値を演算し、この各正常画素の深度修正値により当該各正常画素の画素値を輝度調整して演算される修正正常画素値と、前記対象画素の非可視欠陥深度により前記対象画素の画素値を輝度調整して演算される修正対象画素値との差の絶対値の所定値に対する補数を演算して各正常画素の重みとし、前記一定領域内の正常画素の修正正常画素値の前記重みを用いた重み付き平均値を演算し、その演算結果を平均正常画素値とする点にある。   According to a fourth characteristic configuration of the image processing apparatus according to the present invention, in addition to any of the first to third characteristic configurations, the average normal pixel value calculating unit includes a reference pixel value of the invisible data and the non-visible data. A depth correction value based on non-visible data of each normal pixel in the fixed region is calculated based on a pixel value of a pixel corresponding to each normal pixel of visible data, and each of the normal pixels is calculated based on the depth correction value of each normal pixel. The difference between the corrected normal pixel value calculated by adjusting the brightness of the pixel value of the normal pixel and the correction target pixel value calculated by adjusting the brightness of the pixel value of the target pixel according to the invisible defect depth of the target pixel. The complement of the absolute value for the predetermined value is calculated to obtain the weight of each normal pixel, the weighted average value using the weight of the corrected normal pixel value of the normal pixel in the fixed area is calculated, and the result of the calculation is average normal It is in a point to be a pixel value.

この第4特徴構成によれば、前記各修正正常画素値と前記修正対象画素値との差の絶対値の所定値に対する補数を各正常画素の重みとして重み付き平均値を演算することにより、輝度調整後において対象画素との画素値の差が大きい正常画素の重みを小さくすることになるので、前記輝度調整後の対象画素との比較において他の正常画素よりも可視データに含まれる画像の濃度差が大きい正常画素の前記平均正常画素値に対する影響を抑えることができる。したがって、前記対象画素の周囲に可視データに含まれる画像の濃度が大きく変化する部分がある場合にもそのような画像の影響を最小限に抑え、前記可視欠陥深度を正確に演算するために適した平均正常画素値を演算することができる。   According to the fourth feature configuration, by calculating a weighted average value using a complement of a predetermined absolute value of a difference between each corrected normal pixel value and the correction target pixel value as a weight of each normal pixel, Since the weight of a normal pixel having a large pixel value difference from the target pixel after adjustment is reduced, the density of the image included in the visible data in comparison with the target pixel after the brightness adjustment is compared to the other normal pixels. The influence of normal pixels having a large difference on the average normal pixel value can be suppressed. Therefore, even when there is a part in which the density of the image included in the visible data changes greatly around the target pixel, it is suitable for minimizing the influence of such an image and calculating the visible defect depth accurately. The average normal pixel value can be calculated.

本発明に係る画像処理装置の第5特徴構成は、上記第4の特徴構成に加えて、前記平均正常画素値演算手段が、複数色の可視データに基づいて各正常画素の重みを演算する際に、前記修正正常画素値に代えて各色の修正正常画素値の平均値を用い、前記修正対象画素値に代えて各色の修正対象画素値の平均値を用いる点にある。   According to a fifth characteristic configuration of the image processing apparatus of the invention, in addition to the fourth characteristic configuration, the average normal pixel value calculating unit calculates a weight of each normal pixel based on visible data of a plurality of colors. In addition, instead of the corrected normal pixel value, an average value of corrected normal pixel values of each color is used, and an average value of corrected target pixel values of each color is used instead of the corrected target pixel value.

この第5特徴構成によれば、例えば、R(赤色)、G(緑色)、B(青色)等のように複数色の可視データが取得されている場合に、これらの複数色の可視データに共通する一つの重みを演算するので、平均正常画素値を演算する際における一定領域内の正常画素の修正正常画素値の重みを用いた重み付き平均値をRGBのそれぞれについて演算するための演算処理量を軽減して処理時間を短縮することができる。   According to the fifth feature configuration, when multiple colors of visible data such as R (red), G (green), and B (blue) are acquired, the visible data of these multiple colors Since one common weight is calculated, a calculation process for calculating a weighted average value for each of RGB using the weights of the corrected normal pixel values of normal pixels in a certain region when calculating the average normal pixel value The processing time can be shortened by reducing the amount.

本発明に係る画像処理方法の特徴構成は、画像記録材料に対して可視光及び非可視光を照射して可視データ及び非可視データをそれぞれ取得し、前記非可視データに基づいて検出された前記可視データに含まれる欠陥画素の修正を行う画像処理方法において、複数の前記欠陥画素の一つを対象画素として選択し、前記非可視データの基準画素値と前記非可視データの前記対象画素に対応する画素の画素値とに基づいて前記対象画素の非可視データに基づく欠陥深度を演算し、前記可視データにおける前記対象画素の周囲の一定領域内の正常画素の平均画素値を演算し、前記対象画素の画素値と前記平均正常画素値とに基づいて前記対象画素の可視データに基づく欠陥深度を演算し、前記対象画素を順次選択することによりそれぞれ複数演算される前記非可視欠陥深度及び可視欠陥深度について回帰分析を行って回帰係数と相関係数とを演算し、前記回帰係数及び相関係数に基づいて前記可視データと非可視データとの欠陥深度のずれを補正する補正係数を演算し、前記補正係数と各欠陥画素の非可視欠陥深度とに基づいて各欠陥画素の輝度調整による修正を行う点にある。   The characteristic configuration of the image processing method according to the present invention is to irradiate the image recording material with visible light and invisible light to obtain visible data and invisible data, respectively, and detect the data detected based on the invisible data. In an image processing method for correcting a defective pixel included in visible data, one of the plurality of defective pixels is selected as a target pixel, and corresponds to a reference pixel value of the non-visible data and the target pixel of the non-visible data Calculating a defect depth based on non-visible data of the target pixel based on a pixel value of the target pixel, calculating an average pixel value of normal pixels in a certain region around the target pixel in the visible data, and calculating the target A defect depth based on visible data of the target pixel is calculated based on the pixel value of the pixel and the average normal pixel value, and a plurality of calculations are performed by sequentially selecting the target pixel. A regression analysis is performed on the invisible defect depth and the visible defect depth to calculate a regression coefficient and a correlation coefficient, and the difference in defect depth between the visible data and the invisible data is calculated based on the regression coefficient and the correlation coefficient. A correction coefficient for correcting the defect is calculated, and correction is performed by adjusting the luminance of each defective pixel based on the correction coefficient and the invisible defect depth of each defective pixel.

この特徴構成によれば、前記第1特徴構成による効果と同じく、可視光と非可視光との波長或いは屈折率の相違に起因する、可視データが欠陥により受ける影響と非可視データが欠陥により受ける影響とのずれを正確に補正する輝度調整を行うことが可能となるとともに、欠陥の補正の際の演算処理量を軽減して処理時間を短縮することが可能となる。   According to this feature configuration, as with the effect of the first feature configuration, the visible data is affected by the defect and the invisible data is affected by the defect due to the difference in wavelength or refractive index between visible light and invisible light. It is possible to perform brightness adjustment that accurately corrects a deviation from the influence, and it is possible to reduce the processing amount by reducing the amount of calculation processing at the time of defect correction.

本発明に係る画像処理プログラムの特徴構成は、画像記録材料に対して可視光及び非可視光を照射して可視データ及び非可視データをそれぞれ取得し、前記非可視データに基づいて検出された前記可視データに含まれる欠陥画素の修正を行う画像処理方法において、複数の前記欠陥画素の一つを対象画素として選択し、前記非可視データの基準画素値と前記非可視データの前記対象画素に対応する画素の画素値とに基づいて前記対象画素の非可視データに基づく欠陥深度を演算し、前記可視データにおける前記対象画素の周囲の一定領域内の正常画素の平均画素値を演算し、前記対象画素の画素値と前記平均正常画素値とに基づいて前記対象画素の可視データに基づく欠陥深度を演算し、前記対象画素を順次選択することによりそれぞれ複数演算される前記非可視欠陥深度及び可視欠陥深度について回帰分析を行って回帰係数と相関係数とを演算し、前記回帰係数及び相関係数に基づいて前記非可視データと非可視データとの欠陥深度のずれを補正する補正係数を演算し、前記補正係数と各欠陥画素の非可視欠陥深度とに基づいて各欠陥画素の輝度調整による修正を行う処理をコンピュータに実行させる点にある。   The characteristic configuration of the image processing program according to the present invention is to obtain visible data and invisible data by irradiating an image recording material with visible light and invisible light, respectively, and detect the data detected based on the invisible data In an image processing method for correcting a defective pixel included in visible data, one of the plurality of defective pixels is selected as a target pixel, and corresponds to a reference pixel value of the non-visible data and the target pixel of the non-visible data Calculating a defect depth based on non-visible data of the target pixel based on a pixel value of the target pixel, calculating an average pixel value of normal pixels in a certain region around the target pixel in the visible data, and calculating the target By calculating the defect depth based on the visible data of the target pixel based on the pixel value of the pixel and the average normal pixel value, and sequentially selecting the target pixel, The regression coefficient and the correlation coefficient are calculated by performing regression analysis on the calculated invisible defect depth and the visible defect depth, and the defect between the invisible data and the invisible data is calculated based on the regression coefficient and the correlation coefficient. A correction coefficient for correcting a shift in depth is calculated, and the computer is caused to execute a process of correcting the defective pixel by brightness adjustment based on the correction coefficient and the invisible defect depth of each defective pixel.

この特徴構成によれば、前記第1特徴構成による効果と同じく、可視光と非可視光との波長或いは屈折率の相違に起因する、可視データが欠陥により受ける影響と非可視データが欠陥により受ける影響とのずれを正確に補正する輝度調整を行うことが可能となるとともに、欠陥の補正の際の演算処理量を軽減して処理時間を短縮することが可能となる。   According to this feature configuration, as with the effect of the first feature configuration, the visible data is affected by the defect and the invisible data is affected by the defect due to the difference in wavelength or refractive index between visible light and invisible light. It is possible to perform brightness adjustment that accurately corrects a deviation from the influence, and it is possible to reduce the processing amount by reducing the amount of calculation processing at the time of defect correction.

以下、本発明に係る画像処理装置1を、画像記録材料である写真フィルム2から画像を読み取って印画紙3に記録する画像プリントシステム4に適用した実施形態について図面に基づいて説明する。図1は、本実施形態に係る画像プリントシステム4の外観を示す斜視図であり、図2は、本実施形態に係る画像プリントシステム4の概略構成を示す模式図であり、図3は、本実施形態に係る画像処理装置1の機能を示すブロック図である。なお、本実施形態においては、非可視光として赤外光を用いることとする。   Hereinafter, an embodiment in which an image processing apparatus 1 according to the present invention is applied to an image printing system 4 that reads an image from a photographic film 2 that is an image recording material and records the image on a photographic paper 3 will be described with reference to the drawings. FIG. 1 is a perspective view illustrating an appearance of an image print system 4 according to the present embodiment, FIG. 2 is a schematic diagram illustrating a schematic configuration of the image print system 4 according to the present embodiment, and FIG. It is a block diagram which shows the function of the image processing apparatus 1 which concerns on embodiment. In the present embodiment, infrared light is used as invisible light.

これらの図に示すように、この画像プリントシステム4は、図示しないフィルム現像機によって現像処理された写真フィルム2の撮影画像コマをデジタル画像データとして読み取る画像読取装置5、取得された画像データに画像処理を施してプリントデータを作成する画像処理装置1、及び画像処理装置1からのプリントデータに基づいて露光処理と現像処理とを行って印画紙3に画像を記録する画像記録装置6を備えて構成されている。   As shown in these drawings, the image print system 4 includes an image reading device 5 that reads a photographed image frame of a photographic film 2 developed by a film developing machine (not shown) as digital image data, and an image in the acquired image data. An image processing device 1 that performs processing to create print data, and an image recording device 6 that performs exposure processing and development processing based on print data from the image processing device 1 and records an image on the photographic paper 3 are provided. It is configured.

画像読取装置5は、いわゆるフィルムスキャナであり、主な構成要素としては、図2に示すように、照明光学系7、ズームレンズ等の撮像光学系8、入射してきた光を可視光と赤外光に分けるダイクロイックミラー9、可視光用センサユニット10、赤外光用センサユニット11を備えている。照明光学系7は、光源としてのハロゲンランプ又は発光ダイオードと、その光源からの光を調光するミラートンネルや拡散板などから構成されている。可視光用センサユニット10は、写真フィルム2の3つの基本色成分、本実施形態においては、R(赤色)光、G(緑色)光、B(青色)光からなる可視光画像を検出するためにそれぞれ適合するカラーフィルタを装着した3つのCCDアレイ10aと、これらのCCDアレイ10aによって検出された可視光信号を処理して基本色成分で構成されたR光、G光、及びB光のそれぞれの可視データを生成して画像処理装置1へ転送する可視光用信号処理回路10bを備えている。また、赤外光用センサユニット11は、写真フィルム2に付いている傷等の欠陥の状態を赤外光画像として検出するために、ダイクロイックミラー9から分岐された赤外光のみを受けるように配置されたCCDアレイ11aと、このCCDアレイ11aによって検出された赤外光信号を処理して赤外データを生成して画像処理装置1へ転送する赤外光用信号処理回路11bを備えている。   The image reading device 5 is a so-called film scanner. As shown in FIG. 2, the image reading device 5 includes an illumination optical system 7, an imaging optical system 8 such as a zoom lens, and incident light as visible light and infrared light. A dichroic mirror 9 for dividing light, a visible light sensor unit 10, and an infrared light sensor unit 11 are provided. The illumination optical system 7 includes a halogen lamp or a light emitting diode as a light source, and a mirror tunnel or a diffusion plate for dimming light from the light source. The visible light sensor unit 10 detects a visible light image composed of the three basic color components of the photographic film 2, in this embodiment, R (red) light, G (green) light, and B (blue) light. Three CCD arrays 10a each equipped with a color filter suitable for each of R, G light, and B light that are composed of basic color components by processing visible light signals detected by these CCD arrays 10a. The visible light signal processing circuit 10b for generating visible data and transferring it to the image processing apparatus 1 is provided. The infrared light sensor unit 11 receives only infrared light branched from the dichroic mirror 9 in order to detect the state of defects such as scratches on the photographic film 2 as an infrared light image. The arranged CCD array 11a and the infrared light signal processing circuit 11b for processing the infrared light signal detected by the CCD array 11a to generate infrared data and transferring it to the image processing apparatus 1 are provided. .

このように構成された画像読取装置5では、写真フィルム2の撮影画像コマが所定の読取位置に位置決めされると、撮影画像コマの読取処理が開始されるが、その際撮影画像コマの投影光像は、フィルム搬送機構12による写真フィルム2の副走査方向への送り操作により、複数のスリット画像に分割された形で順次可視光用センサユニット10及び赤外光用センサユニット11によって読み取られ、R、G、Bの各色成分の可視光信号並びに赤外成分の赤外光信号に光電変換され、デジタル画像データである可視データ及び赤外データとして画像処理装置1に送られる。このような、照明光学系7、撮像光学系8、可視光用センサユニット10及び赤外光用センサユニット11の各制御は画像処理装置1によって行われており、本実施形態では、画像処理装置1の一部の機能部分が画像読取装置5の構成要素となっている。   In the image reading device 5 configured as described above, when the photographed image frame of the photographic film 2 is positioned at a predetermined reading position, reading processing of the photographed image frame is started. The image is sequentially read by the visible light sensor unit 10 and the infrared light sensor unit 11 in a form divided into a plurality of slit images by a feeding operation of the photographic film 2 in the sub-scanning direction by the film transport mechanism 12. It is photoelectrically converted into a visible light signal of each color component of R, G, and B and an infrared light signal of an infrared component, and sent to the image processing apparatus 1 as visible data and infrared data that are digital image data. Each control of the illumination optical system 7, the imaging optical system 8, the visible light sensor unit 10, and the infrared light sensor unit 11 is performed by the image processing apparatus 1. In the present embodiment, the image processing apparatus A part of the functional part 1 is a component of the image reading device 5.

画像処理装置1は、ここでは基本的には汎用パソコンから構成されており、更に、この画像プリントシステム4の操作画面を表示するモニタ13、デジタルカメラ等のメモリカード等から画像を読み込むメディアリーダ14、オペレータによる操作入力に用いられるキーボード15及びマウス16等が付属して構成されている。   Here, the image processing apparatus 1 is basically composed of a general-purpose personal computer, and further includes a monitor 13 for displaying an operation screen of the image print system 4 and a media reader 14 for reading an image from a memory card such as a digital camera. A keyboard 15 and a mouse 16 used for operation input by the operator are attached.

画像処理装置1は、CPU17を中核部材として、入力されたデータに対して種々の処理を行うための機能部がハードウエア又はソフトウエア或いはその両方で実装されているが、本発明に特に関係する機能部としては、図3に示すように、可視データ及び赤外データを各種処理のために一時的に格納するメモリ18、メモリ18に格納されている可視データ及び赤外データを用いて欠陥画素の検出及び修正を行う欠陥修正部19と、メモリ18に展開されている可視データに対して色調補正やフィルタリング(ぼかしやシャープネスなど)やトリミング等の欠陥修正以外の各種画像処理を施す画像調整部20、画像データやその他の表示アイテムをビデオメモリに取り込むとともにこのビデオメモリに展開されたイメージをビデオコントローラによってビデオ信号に変換してモニタ13に送るビデオ制御部21、欠陥修正部19及び画像調整部20で処理された最終的な画像データをプリントデータに変換して画像記録装置6の露光プリント部22に転送するプリントデータ生成部23、グラフィカルユーザーインターフェイス(GUI)を用いて作り出された操作画面の下でキーボード15及びマウス16等を通じて入力された操作指令や予めプログラム化された操作指令に基づいて各機能部を制御するプリント管理部24が挙げられる。   The image processing apparatus 1 has a CPU 17 as a core member, and a functional unit for performing various processes on input data is implemented by hardware and / or software, and is particularly related to the present invention. As shown in FIG. 3, the functional unit includes a memory 18 for temporarily storing visible data and infrared data for various processing, and a defective pixel using the visible data and infrared data stored in the memory 18. A defect correction unit 19 that detects and corrects the image, and an image adjustment unit that performs various image processes other than defect correction such as color tone correction, filtering (blurring, sharpness, etc.) and trimming on the visible data developed in the memory 18 20. The image data and other display items are taken into the video memory, and the image developed in the video memory is converted into the video controller. The final image data processed by the video control unit 21, the defect correction unit 19, and the image adjustment unit 20, which is converted into a video signal by the camera and sent to the monitor 13, is converted into print data and the exposure print unit of the image recording apparatus 6. 22 based on an operation command input through the keyboard 15 and the mouse 16 under the operation screen created using the print data generation unit 23 and the graphical user interface (GUI), or a pre-programmed operation command. There is a print management unit 24 that controls each functional unit.

図4(b)は、赤外データ及び可視データの両方を含む画像データ中のある一つの直線Lに沿って配列された画素の画素値を、赤外データ及びR光、G光、及びB光のそれぞれの可視データについて表したグラフであり、この直線Lは、図4(a)に示すように画像データの欠陥部Pを通り、この欠陥部Pの中に対象画素Oが選択されている。
欠陥修正部19は、この図4に示すように、赤外データにおける対象画素Oの画素値と周囲の正常画素の画素値との差に基づいて赤外欠陥深度Dfm,nを演算するとともに、R光、G光、及びB光のそれぞれの可視データにおける対象画素Oの画素値と周囲の正常画素の画素値との差に基づいて可視欠陥深度Rfm,n、Gfm,n、Bfm,nを演算し、これらに基づいて可視光と非可視光との波長或いは屈折率の相違に起因する、可視データが欠陥により受ける影響と非可視データが欠陥により受ける影響とのずれを補正する補正係数を演算し、この補正係数Wと各欠陥画素の赤外欠陥深度Dfm,nとに基づいて各欠陥画素の輝度調整による修正を行うものである。
FIG. 4B shows pixel values of pixels arranged along one straight line L in image data including both infrared data and visible data, and the infrared data, R light, G light, and B FIG. 4 is a graph showing each visible data of light, and this straight line L passes through a defective portion P of image data as shown in FIG. 4A, and the target pixel O is selected in the defective portion P. Yes.
As shown in FIG. 4, the defect correcting unit 19 calculates the infrared defect depth Df m, n based on the difference between the pixel value of the target pixel O in the infrared data and the pixel values of surrounding normal pixels. , Rf m, n , Gf m, n , Bf based on the difference between the pixel value of the target pixel O and the pixel values of the surrounding normal pixels in the visible data of R light, G light, and B light Calculate m and n , and based on these, correct the difference between the effect of visible data due to defects and the effect of non-visible data due to defects due to the difference in wavelength or refractive index between visible light and invisible light. The correction coefficient to be calculated is calculated, and correction is performed by adjusting the luminance of each defective pixel based on the correction coefficient W and the infrared defect depth Dfm , n of each defective pixel.

この欠陥修正部19は、メモリ18に格納された赤外データに基づいて、R光、G光、及びB光の可視データに含まれる各画素から欠陥画素を検出し、当該検出された欠陥画素の各座標を登録した欠陥画素マップを作成する欠陥画素検出部25と、欠陥画素マップに登録された複数の欠陥画素の一つを対象画素Oとして選択する対象画素選択部26と、赤外データの基準画素値CFと赤外データの対象画素Oに対応する画素の画素値IRdatm,nとに基づいて対象画素Oの赤外データに基づく欠陥深度である赤外欠陥深度Dfm,nを演算する赤外欠陥深度演算部27と、R光、G光、及びB光のそれぞれの可視データにおける対象画素Oの周囲の一定領域E内の正常画素の平均画素値である平均正常画素値Ravem,n、Gavem,n、Bavem,nを演算する平均正常画素値演算部28と、対象画素Oの画素値Rdatm,n、Gdatm,n、Bdatm,nと平均正常画素値Ravem,n、Gavem,n、Bavem,nとに基づいて対象画素Oの可視データに基づく欠陥深度である可視欠陥深度Rfm,n、Gfm,n、Bfm,nを演算する可視欠陥深度演算部29と、対象画素Oを順次選択することにより赤外欠陥深度演算部27及び可視欠陥深度演算部29においてそれぞれ複数演算される赤外欠陥深度Dfm,n及び可視欠陥深度Rfm,n、Gfm,n、Bfm,nを記憶する演算結果記憶部30と、この演算結果記憶部30に記憶された複数の赤外欠陥深度Dfm,nと可視欠陥深度Rfm,n、Gfm,n、Bfm,nとについて回帰分析を行い、回帰係数Raと相関係数Rrとを演算する回帰演算部31と、回帰係数Ra及び相関係数Rrに基づいて可視データと非可視データとの欠陥深度(すなわち欠陥による画素値の低減の程度)のずれを補正する補正係数Wを演算する補正係数演算部32と、この補正係数Wと各欠陥画素の赤外欠陥深度Dfm,nとに基づいて各欠陥画素の輝度調整による修正を行う修正部33とを有している。 The defect correcting unit 19 detects a defective pixel from each pixel included in the visible data of the R light, the G light, and the B light based on the infrared data stored in the memory 18, and the detected defective pixel A defective pixel detection unit 25 that creates a defective pixel map in which each coordinate is registered, a target pixel selection unit 26 that selects one of a plurality of defective pixels registered in the defective pixel map as a target pixel O, and infrared data Infrared defect depth Df m, n that is a defect depth based on the infrared data of the target pixel O based on the reference pixel value CF and the pixel value IRdat m, n of the pixel corresponding to the target pixel O of the infrared data An infrared defect depth calculation unit 27 to calculate, and an average normal pixel value Rave that is an average pixel value of normal pixels in a fixed region E around the target pixel O in each visible data of R light, G light, and B light Average normal pixel value for calculating m, n , Gave m, n and Bave m, n The target pixel based on the arithmetic unit 28 and the pixel values Rdat m, n , Gdat m, n and Bdat m, n of the target pixel O and the average normal pixel values Rave m, n , Gave m, n and Bave m, n O visible defect depth Rf m is the defect depth based on visual data, n, infrared defects by Gf m, n, Bf m, visible defect depth calculation unit 29 for calculating the n, sequentially selects a target pixel O infrared defect depth Df m, n and visible defects depth Rf m being more operations, respectively, in depth calculation unit 27 and the visible defect depth calculation unit 29, n, Gf m, n , Bf m, the operation result storing n storage unit perform a 30, a plurality of infrared defect depth Df m stored in the calculation result storage unit 30, n and visible defects depth Rf m, n, Gf m, n, Bf m, for the n regression analysis, regression coefficients Regression calculation unit 31 for calculating Ra and correlation coefficient Rr, and the defect depth of visible data and non-visible data based on regression coefficient Ra and correlation coefficient Rr (that is, the pixel value due to the defect) A correction coefficient calculation unit 32 for calculating a correction coefficient W for correcting the deviation of the reduced degree of), the correction coefficient W and each defective pixel infrared defect depth Df m, according to the brightness adjustment of each defective pixel based on the n And a correction unit 33 that performs correction.

ここで、本実施形態における対象画素選択部26、赤外欠陥深度演算部27、平均正常画素値演算部28、可視欠陥深度演算部29、演算結果記憶部30、回帰演算部31、補正係数演算部32、及び修正部33は、それぞれ特許請求の範囲における対象画素選択手段、非可視欠陥深度演算手段、平均正常画素値演算手段、可視欠陥深度演算手段、記憶手段、回帰演算手段、補正係数演算手段、及び修正手段に相当する。なお、この欠陥修正部19における欠陥画素の修正処理については、後に詳細に説明する。   Here, the target pixel selection unit 26, the infrared defect depth calculation unit 27, the average normal pixel value calculation unit 28, the visible defect depth calculation unit 29, the calculation result storage unit 30, the regression calculation unit 31, and the correction coefficient calculation in the present embodiment. The unit 32 and the correction unit 33 are respectively a target pixel selection unit, a non-visible defect depth calculation unit, an average normal pixel value calculation unit, a visible defect depth calculation unit, a storage unit, a regression calculation unit, and a correction coefficient calculation. It corresponds to a means and a correction means. The defective pixel correcting process in the defect correcting unit 19 will be described in detail later.

画像記録装置6は、図2に示されているように、2つの印画紙マガジン34に納められたロール状の印画紙3を引き出してシートカッター35でプリントサイズに切断すると共に、このように切断された印画紙3に対し、バックプリント部36で色補正情報やコマ番号などのプリント処理情報を印画紙3の裏面に印字するとともに、露光プリント部22で印画紙3の表面に撮影画像の露光を行い、この露光後の印画紙3を複数の現像処理槽を有した処理槽ユニット37に送り込んで現像処理する。乾燥の後に装置上部の横送りコンベア38からソータ39に送られた印画紙3は、このソータ39の複数のトレイ40にオーダ単位で仕分けられた状態で集積される(図1参照)。   As shown in FIG. 2, the image recording apparatus 6 pulls out the roll-shaped photographic paper 3 stored in the two photographic paper magazines 34 and cuts it into a print size by the sheet cutter 35. Print processing information such as color correction information and frame number is printed on the back surface of the photographic paper 3 by the back print unit 36 on the printed paper 3 and the exposure print unit 22 exposes the photographed image on the surface of the photographic paper 3. The exposed photographic paper 3 is sent to a processing tank unit 37 having a plurality of development processing tanks for development processing. After drying, the photographic paper 3 sent to the sorter 39 from the transverse feed conveyor 38 at the top of the apparatus is collected in a state of being sorted in units of orders on a plurality of trays 40 of the sorter 39 (see FIG. 1).

また、画像記録装置6には、印画紙3に対する各種処理に合わせた搬送速度で印画紙3を搬送するために印画紙搬送機構41が敷設されている。印画紙搬送機構41は、印画紙搬送方向に関して露光プリント部22の前後に配置されたチャッカー式印画紙搬送ユニット41aを含む複数の挟持搬送ローラ対から構成されている。露光プリント部22には、副走査方向に搬送される印画紙3に対して、主走査方向に沿って画像記録装置6からのプリントデータに基づいてR、G、Bの3原色のレーザ光線の照射を行うライン露光ヘッドが設けられている。処理槽ユニット37は、発色現像処理液を貯留する発色現像槽37aと、漂白定着処理液を貯留する漂白定着槽37bと、安定処理液を貯留する安定槽37cとを備えている。   In addition, the image recording apparatus 6 is provided with a photographic paper transport mechanism 41 for transporting the photographic paper 3 at a transport speed in accordance with various processes for the photographic paper 3. The photographic paper transport mechanism 41 is composed of a plurality of nipping and transporting roller pairs including a chucker type photographic paper transport unit 41a disposed before and after the exposure print unit 22 in the photographic paper transport direction. The exposure print unit 22 applies the laser beams of the three primary colors R, G, and B to the photographic paper 3 conveyed in the sub-scanning direction based on the print data from the image recording device 6 along the main scanning direction. A line exposure head for irradiating is provided. The processing tank unit 37 includes a color developing tank 37a for storing a color developing processing liquid, a bleach-fixing tank 37b for storing a bleach-fixing processing liquid, and a stabilizing tank 37c for storing a stable processing liquid.

次に、本実施形態に係る画像処理装置1における欠陥画素の修正処理について、図5に示すフローチャートに従って詳細に説明する。   Next, the defective pixel correction processing in the image processing apparatus 1 according to the present embodiment will be described in detail according to the flowchart shown in FIG.

まず、画像読取装置5の可視光用センサユニット10及び赤外光用センサユニット11によって取得された可視データ及び赤外データをメモリ18に取り込む(#01)。そして、欠陥画素検出部25において、#01の処理でメモリ18に格納された赤外データに基づいて、R光、G光、及びB光の可視データに含まれる各画素から欠陥画素を検出し、当該検出された欠陥画素の各座標を登録した欠陥画素マップを作成する(#02)。この処理は、赤外データに含まれる各画素の画素値が、予め設定した一定の閾値CF2以下である場合に欠陥画素として検出し、欠陥画素マップにその座標を登録することにより行うことができる。ここで用いる一定の閾値CF2は、赤外データの基準画素値CFに基づいて定める画素値とし、この基準画素値CFよりもやや低い画素値に設定すると好適である。ここで、赤外データの基準画素値CFとしては、写真フィルム2に傷や埃等の欠陥が存在しない理想的な状態における赤外データの画素値の平均値を、赤外データに基づいて演算した値を用いると好適である。   First, the visible data and infrared data acquired by the visible light sensor unit 10 and the infrared light sensor unit 11 of the image reading device 5 are taken into the memory 18 (# 01). Then, the defective pixel detection unit 25 detects a defective pixel from each pixel included in the visible data of the R light, the G light, and the B light based on the infrared data stored in the memory 18 in the process of # 01. Then, a defective pixel map in which the coordinates of the detected defective pixels are registered is created (# 02). This process can be performed by detecting a defective pixel when the pixel value of each pixel included in the infrared data is equal to or less than a predetermined threshold CF2 and registering the coordinates in the defective pixel map. . The constant threshold CF2 used here is preferably a pixel value determined based on the reference pixel value CF of infrared data, and is preferably set to a pixel value slightly lower than the reference pixel value CF. Here, as the reference pixel value CF of the infrared data, an average value of the pixel values of the infrared data in an ideal state where the photographic film 2 is free from defects such as scratches and dust is calculated based on the infrared data. It is preferable to use the values obtained.

次に、対象画素選択部26において、#02の処理で検出されて欠陥画素マップに登録された複数の欠陥画素の中から一つの欠陥画素を対象画素Oとして選択する(#03)。ここでは、選択された対象画素Oの座標を(m,n)とする。本実施形態においては、処理対象の画像領域はNX×NY画素の大きさを有することとし、mは1〜NX、nは1〜NYの値をとり得る。この対象画素Oは、欠陥画素マップに登録された複数の欠陥画素の中から一つずつ順次選択され、最終的には、処理対象の画像領域の全ての欠陥画素が対象画素Oとして選択されることになる。ここで、処理対象の画像領域として、本実施形態においては、写真フィルム2に記録された一枚の画像を対象とする例について説明するが、一枚の画像の中の所定の範囲を処理対象の画像領域とすることも可能であるし、また、写真フィルム2の全体を一つの処理対象の画像領域とすることも可能である。   Next, the target pixel selection unit 26 selects one defective pixel as the target pixel O from the plurality of defective pixels detected in the process of # 02 and registered in the defective pixel map (# 03). Here, the coordinates of the selected target pixel O are (m, n). In the present embodiment, it is assumed that the image area to be processed has a size of NX × NY pixels, and m can take a value of 1 to NX and n can take a value of 1 to NY. The target pixels O are sequentially selected one by one from a plurality of defective pixels registered in the defective pixel map, and finally all defective pixels in the image area to be processed are selected as the target pixels O. It will be. Here, as an image area to be processed, in this embodiment, an example in which one image recorded on the photographic film 2 is an object will be described. However, a predetermined range in one image is to be processed. The entire photographic film 2 can be used as one image area to be processed.

そして、赤外欠陥深度演算部27において、赤外データの基準画素値CFと#03の処理で選択された対象画素Oに対応する赤外データの画素の画素値IRdatm,nとに基づいて対象画素Oの赤外データに基づく欠陥深度である赤外欠陥深度Dfm,nを演算する(#04)。具体的には、以下の式(1)に従って、赤外データの基準画素値CFと赤外データの対象画素Oに対応する画素の画素値IRdatm,nとの差を演算し、その演算結果を赤外欠陥深度Dfm,nとする。 Then, in the infrared defect depth calculation unit 27, based on the reference pixel value CF of infrared data and the pixel value IRdat m, n of the pixel of infrared data corresponding to the target pixel O selected in the process of # 03. An infrared defect depth Dfm , n , which is a defect depth based on the infrared data of the target pixel O, is calculated (# 04). Specifically, the difference between the reference pixel value CF of the infrared data and the pixel value IRdat m, n of the pixel corresponding to the target pixel O of the infrared data is calculated according to the following formula (1), and the calculation result Is the infrared defect depth Dfm , n .

Figure 2005142621
Figure 2005142621

ここで、赤外欠陥深度Dfm,nは、図4(b)に示すように、赤外データの対象画素Oに対応する画素の画素値IRdatm,nが、欠陥が存在しない理想的な状態における赤外データの画素値を表す赤外データの基準画素値CFに対してどれだけ低減しているか、すなわち傷や埃等の欠陥による赤外データの画素値の低減の程度を表す値である。 Here, as shown in FIG. 4B, the infrared defect depth Df m, n is ideal when the pixel value IRdat m, n of the pixel corresponding to the target pixel O of infrared data does not have a defect. This is a value indicating how much the infrared data reference pixel value CF represents the pixel value of the infrared data in the state, that is, the degree of reduction of the pixel value of the infrared data due to defects such as scratches and dust. is there.

次に、平均正常画素値演算部28において、R光、G光、及びB光のそれぞれの可視データにおける対象画素Oの周囲の一定領域E内の正常画素の平均画素値である平均正常画素値Ravem,n、Gavem,n、Bavem,nを演算する(#05)。図6は、対象画素Oとその周囲の一定領域の一例を示す説明図であり、図7は、この平均正常画素値演算部28における平均正常画素値Ravem,n、Gavem,n、Bavem,nの演算処理を示すフローチャートである。図6に示すように、本実施形態では一定領域Eは、5×5画素の正方形領域としている。この平均正常画素値Ravem,n、Gavem,n、Bavem,nの演算は、具体的には以下のようにして行う。 Next, in the average normal pixel value calculation unit 28, an average normal pixel value that is an average pixel value of normal pixels in the fixed region E around the target pixel O in the visible data of each of R light, G light, and B light. Rave m, n , Gave m, n and Bave m, n are calculated (# 05). FIG. 6 is an explanatory diagram showing an example of the target pixel O and a fixed area around it. FIG. 7 shows average normal pixel values Rave m, n , Gave m, n , Bave in the average normal pixel value calculation unit 28. It is a flowchart which shows the calculation process of m, n . As shown in FIG. 6, in this embodiment, the fixed area E is a square area of 5 × 5 pixels. The average normal pixel values Rave m, n , Gave m, n , and Bave m, n are specifically calculated as follows.

まず、対象画素Oの周囲の一定領域E内の各画素の内、欠陥画素マップに登録されている欠陥画素を除く各正常画素について、赤外データの基準画素値CFと赤外データの各正常画素に対応する画素の画素値IRdatm+i,n+jとに基づいて深度修正値Dfi,jを演算する(#21)。具体的には、以下の式(2)に従って、赤外欠陥深度Dfm,nの演算と同様に、赤外データの基準画素値CFと赤外データの各正常画素に対応する画素の画素値IRdatm+i,n+jとの差を演算する。 First, for each normal pixel excluding the defective pixel registered in the defective pixel map among the pixels in the fixed region E around the target pixel O, the normal pixel value CF of the infrared data and each normal value of the infrared data The depth correction value Dfi , j is calculated based on the pixel value IRdat m + i, n + j of the pixel corresponding to the pixel (# 21). Specifically, according to the following formula (2), the pixel value of the pixel corresponding to each normal pixel of the infrared data reference pixel value CF and the infrared data, as in the calculation of the infrared defect depth Dfm , n. Calculate the difference from IRdat m + i, n + j .

Figure 2005142621
Figure 2005142621

ここで、(i,j)は対象画素Oの座標(m,n)からのx方向及びy方向のそれぞれについての画素数で表される距離であり、対象画素Oの周囲の画素の座標は(m+i,n+j)という座標で表される。本実施形態においては、一定領域Eは5×5画素の正方形領域であるので、i及びjは、それぞれ−2から2の値をとる。   Here, (i, j) is a distance represented by the number of pixels in the x direction and y direction from the coordinates (m, n) of the target pixel O, and the coordinates of the pixels around the target pixel O are It is represented by coordinates (m + i, n + j). In the present embodiment, since the fixed area E is a square area of 5 × 5 pixels, i and j take values from −2 to 2, respectively.

上記のとおり正常画素と欠陥画素との判別は、基準画素値CFよりもやや低い画素値に設定された閾値CF2を用いて行われていることから、正常画素と判断された画素であっても基準画素値CFを基準とすればその画素値は若干低いものとなっているはずである。そこで、R光、G光、及びB光のそれぞれの可視データについての前記一定領域E内の各正常画素の画素値Rdatm+i,n+j、Gdatm+i,n+j、Bdatm+i,n+jを深度修正値Dfi,jにより輝度調整して修正正常画素値Rdat´i,j、Gdat´i,j、Bdat´i,jを演算する(#22)。具体的には、以下の式(3)〜(5)に従って、各正常画素の画素値Rdatm+i,n+j、Gdatm+i,n+j、Bdatm+i,n+jに深度修正値Dfi,jをそれぞれ加算して修正正常画素値Rdat´i,j、Gdat´i,j、Bdat´i,jを演算する。 As described above, the discrimination between the normal pixel and the defective pixel is performed using the threshold value CF2 set to a pixel value slightly lower than the reference pixel value CF. If the reference pixel value CF is used as a reference, the pixel value should be slightly lower. Therefore, the pixel values Rdat m + i, n + j , Gdat m + i, n + j , Bdat m of the normal pixels in the fixed region E for the visible data of R light, G light, and B light respectively. The brightness of + i, n + j is adjusted by the depth correction value Df i, j to calculate corrected normal pixel values Rdat ′ i, j , Gdat ′ i, j , Bdat ′ i, j (# 22). Specifically, according to the following formulas (3) to (5), the pixel values Rdat m + i, n + j , Gdat m + i, n + j , Bdat m + i, n + j of each normal pixel are set. The corrected normal pixel values Rdat ′ i, j , Gdat ′ i, j , and Bdat ′ i, j are calculated by adding the depth correction values Df i, j respectively.

Figure 2005142621
Figure 2005142621

次に、これらの可視データの修正正常画素値Rdat´i,j、Gdat´i,j、Bdat´i,jのRGB平均値Ddat´i,jを演算する(#23)。具体的には、以下の式(6)に従って修正正常画素値Rdat´i,j、Gdat´i,j、Bdat´i,jの単純平均値を演算する。 Next, the RGB average values Ddat ′ i, j of the corrected normal pixel values Rdat ′ i, j , Gdat ′ i, j and Bdat ′ i, j of these visible data are calculated (# 23). Specifically, a simple average value of the corrected normal pixel values Rdat ′ i, j , Gdat ′ i, j and Bdat ′ i, j is calculated according to the following equation (6).

Figure 2005142621
Figure 2005142621

また、R光、G光、及びB光のそれぞれの可視データについての前記対象画素Oの画素値Rdatm,n、Gdatm,n、Bdatm,nを赤外欠陥深度Dfm,nにより輝度調整した修正対象画素値Rdat´m,n、Gdat´m,n、Bdat´m,nのRGB平均値Ddat´m,nを演算する(#24)。具体的には、以下の式(7)に従って演算する。 In addition, the pixel values Rdat m, n , Gdat m, n , and Bdat m, n of the target pixel O for the visible data of R light, G light, and B light are represented by the brightness of the infrared defect depth Df m, n. adjusted correction target pixel value Rdat' m, n, Gdat' m, n, Bdat' m, RGB average value of n Ddat' m, calculates a n (# 24). Specifically, the calculation is performed according to the following equation (7).

Figure 2005142621
Figure 2005142621

そして、修正正常画素値Rdat´i,j、Gdat´i,j、Bdat´i,jのRGB平均値Ddat´i,jと、修正対象画素値Rdat´m,n、Gdat´m,n、Bdat´m,nのRGB平均値Ddat´m,nとの差の絶対値の所定値αに対する補数を演算し、これを各正常画素のRGB共通の重みWgt i,jとする(#25)。この重みWgt i,jは、具体的には、以下の式(8)に従って演算する。 The modified normal pixel value Rdat' i, j, Gdat' i, j, Bdat' i, RGB average values Ddat' i of j, j and, corrected target pixel value Rdat' m, n, Gdat' m, n, Bdat' m, RGB average value of n Ddat' m, calculates the complement for a given value α of the absolute value of the difference between n, which is referred to as RGB common weight Wgt i, j of each normal pixel (# 25) . Specifically , the weight Wgt i, j is calculated according to the following equation (8).

Figure 2005142621
Figure 2005142621

ここで、重みWgt i,jは、対象画素Oとその周囲の各正常画素との間で、可視データの画像の濃度に大きな差がある場合に、そのような濃度差のある正常画素の平均正常画素値Ravem,n、Gavem,n、Bavem,nに対する影響を少なくすることにより、可視データの画像に含まれる絵柄の影響を抑えるための係数である。したがって、所定値αの値は、対象画素Oと正常画素との濃度差を示す修正正常画素値Rdat´i,j、Gdat´i,j、Bdat´i,jのRGB平均値Ddat´i,jと修正対象画素値Rdat´m,n、Gdat´m,n、Bdat´m,nのRGB平均値Ddat´m,nとの差が、平均正常画素値Ravem,n、Gavem,n、Bavem,nとして用いるのに不適当である限界の値に設定する。このような所定値αの値は実験により求めることが可能であるが、例えば、可視データの各画素がRGB毎に12bitのデータ量を有して画素値が0〜4095までの値をとり、各画素値の自然対数を用いて上記式(8)の演算を行う場合であれば、「α=1」程度とすると好適である。そして、修正正常画素値Rdat´i,j、Gdat´i,j、Bdat´i,jのRGB平均値Ddat´i,jと修正対象画素値Rdat´m,n、Gdat´m,n、Bdat´m,nのRGB平均値Ddat´m,nとの差の絶対値が所定値αより大きい場合、すなわち重みWgt i,j<0のときは、重みWgt i,j=0とし、そのような濃度差のある正常画素の平均正常画素値Ravem,n、Gavem,n、Bavem,nに対する影響をゼロとする。 Here, the weight Wgt i, j is the average of the normal pixels having such a density difference when there is a large difference in the density of the image of the visible data between the target pixel O and the surrounding normal pixels. This is a coefficient for suppressing the influence of the pattern included in the image of the visible data by reducing the influence on the normal pixel values Rave m, n , Gave m, n and Bave m, n . Therefore, the value of the predetermined value α is the RGB average value Ddat ′ i, J of the corrected normal pixel values Rdat ′ i, j , Gdat ′ i, j , Bdat ′ i, j indicating the density difference between the target pixel O and the normal pixel . j and correction target pixel value Rdat' m, n, Gdat' m, n, Bdat' m, RGB average value of n Ddat' m, the difference between n has an average normal pixel value Rave m, n, Gave m, n , Bave m, n is set to a limit value that is inappropriate for use. The value of the predetermined value α can be obtained by experiments. For example, each pixel of visible data has a data amount of 12 bits for each RGB, and the pixel value takes a value from 0 to 4095. In the case of performing the calculation of the above formula (8) using the natural logarithm of each pixel value, it is preferable to set “α = 1” or so. Then, the RGB average values Ddat ′ i, j of the corrected normal pixel values Rdat ′ i, j , Gdat ′ i, j , Bdat ′ i, j and the correction target pixel values Rdat ′ m, n , Gdat ′ m, n , Bdat 'm, if RGB average value of n Ddat' m, the absolute value of the difference between the n is greater than the predetermined value alpha, i.e. when the weight Wgt i, j <0, the weight Wgt i, j = 0, so The influence on the average normal pixel values Rave m, n , Gave m, n and Bave m, n of normal pixels having a large density difference is set to zero.

その後、一定領域E内の正常画素の修正正常画素値Rdat´i,j、Gdat´i,j、Bdat´i,jの重みWgt i,jを用いた重み付き平均値をRGBのそれぞれについて演算し、その演算結果を平均正常画素値Ravem,n、Gavem,n、Bavem,nとする(#26)。具体的には、以下の式(9)〜(11)に従って演算する。 After that, the weighted average value using the weight Wgt i, j of the corrected normal pixel values Rdat ′ i, j , Gdat ′ i, j , and Bdat ′ i, j of the normal pixels in the fixed region E is calculated for each of RGB. The calculation results are set as average normal pixel values Rave m, n , Gave m, n and Bave m, n (# 26). Specifically, the calculation is performed according to the following equations (9) to (11).

Figure 2005142621
Figure 2005142621

以上で平均正常画素値Ravem,n、Gavem,n、Bavem,nの演算処理(#05)を終了する。なお、平均正常画素値Ravem,n、Gavem,n、Bavem,nの演算処理方法は、上記に限定されるものではなく、各正常画素のRGB共通の重みWgt i,jをすべて1とし、修正正常画素値Rdat´i,j、Gdat´i,j、Bdat´i,jの単純平均値を平均正常画素値Ravem,n、Gavem,n、Bavem,nとすることも可能である。また、重みWgt i,jの演算に際して、RGB毎に、修正正常画素値Rdat´i,j、Gdat´i,j、Bdat´i,jと、修正対象画素値Rdat´m,n、Gdat´m,n、Bdat´m,nとの差の絶対値の所定値αに対する補数を演算し、これらを各正常画素のRGB毎の重みWgtr i,j、Wgtg i,j、Wgtbi,jとすることも可能である。 The average normal pixel values Rave m, n , Gave m, n and Bave m, n calculation processing (# 05) is thus completed. Note that the calculation processing method of the average normal pixel values Rave m, n , Gave m, n , and Bave m, n is not limited to the above, and all RGB common weights Wgt i, j of each normal pixel are 1 The simple average value of the corrected normal pixel values Rdat ′ i, j , Gdat ′ i, j , Bdat ′ i, j may be set as the average normal pixel values Rave m, n , Gave m, n , Bave m, n Is possible. When calculating the weight Wgt i, j , the corrected normal pixel values Rdat ′ i, j , Gdat ′ i, j , Bdat ′ i, j and the correction target pixel values Rdat ′ m, n , Gdat ′ are calculated for each RGB. The complement of the absolute value of the difference between m, n and Bdat ′ m, n with respect to a predetermined value α is calculated, and these are calculated as weights Wgtr i, j , Wgtg i, j , Wgtb i, j for each normal pixel RGB. It is also possible to do.

次に、図5に示すように、対象画素Oの周囲の一定領域E内の正常画素の有効度の和が一定の有効閾値β以上か否かについて判断する(#06)。ここで、正常画素の有効度は、対象画素Oとその周囲の各正常画素との間における可視データの画像の濃度差を考慮し、この濃度差が小さい程有効な正常画素と判断するための値であり、本実施形態においては、各正常画素の有効度として上記各正常画素のRGB共通の重みWgt i,jを用いる。また、有効閾値βとしては、5×5画素の一定領域E内の全ての正常画素の重みWgt i,jの和の最大値「24α」の半分の値「12α」を用いる。そして、一定領域E内の正常画素の有効度すなわち重みWgt i,jの和が一定の有効閾値β(=12α)未満である場合には(#06:NO)、その対象画素Oの周囲の一定領域E内に有効な正常画素が少ないと判断して、その対象画素Oを以降の処理の対象から除外することとし、処理は#03へ戻り次の対象画素Oを選択する。一方、一定領域E内の正常画素の有効度すなわち重みWgt i,jの和が一定の有効閾値β(=12α)以上である場合には(#06:YES)、その対象画素Oについての処理を継続することとし、処理は#07へ進む。 Next, as shown in FIG. 5, it is determined whether or not the sum of the validity levels of normal pixels in a certain area E around the target pixel O is equal to or greater than a certain effective threshold value β (# 06). Here, the effectiveness of the normal pixel is determined in consideration of the density difference of the image of the visible data between the target pixel O and each of the surrounding normal pixels, and the smaller the density difference, the more effective the normal pixel. In this embodiment, the weight Wgt i, j common to RGB of each normal pixel is used as the effectiveness of each normal pixel. Further, as the effective threshold value β, a value “12α” which is a half of the maximum value “24α” of the sum of the weights Wgt i, j of all the normal pixels in the fixed region E of 5 × 5 pixels is used. If the effectiveness of normal pixels in the fixed region E, that is , the sum of the weights Wgt i, j is less than a certain effective threshold β (= 12α) (# 06: NO), the area around the target pixel O When it is determined that there are few valid normal pixels in the certain area E, the target pixel O is excluded from the subsequent processing targets, and the process returns to # 03 to select the next target pixel O. On the other hand, when the effectiveness of normal pixels in the fixed region E, that is , the sum of the weights Wgt i, j is equal to or greater than a certain effective threshold β (= 12α) (# 06: YES), processing for the target pixel O The process proceeds to # 07.

その後、可視欠陥深度演算部29において、対象画素Oの画素値datm,n、Gdatm,n、Bdatm,nと平均正常画素値Ravem,n、Gavem,n、Bavem,nとに基づいて対象画素Oの可視データに基づく欠陥深度である可視欠陥深度Rfm,n、Gfm,n、Bfm,nを演算する(#07)。具体的には、以下の式(12)〜(14)に従って、平均正常画素値Ravem,n、Gavem,n、Bavem,nと対象画素Oの画素値datm,n、Gdatm,n、Bdatm,nとの差を演算する。 Thereafter, in the visible defect depth calculation unit 29, the pixel values dat m, n , Gdat m, n and Bdat m, n of the target pixel O and the average normal pixel values Rave m, n , Gave m, n and Bave m, n are obtained. calculating a visible defect depth Rf m is the defect depth based on visual data of the target pixel O, n, Gf m, n , Bf m, the n based on (# 07). Specifically, according to the following formulas (12) to (14), the average normal pixel values Rave m, n , Gave m, n , Bave m, n and the pixel values dat m, n , Gdat m, n, calculates the difference between Bdat m, n.

Figure 2005142621
Figure 2005142621

この可視欠陥深度Rfm,n、Gfm,n、Bfm,nは、図4(b)に示すように、R光、G光、及びB光のそれぞれの可視データにおける対象画素Oの画素値が、対象画素Oの周囲の正常画素の平均値である平均正常画素値Ravem,n、Gavem,n、Bavem,nに対してどれだけ低減しているか、すなわち傷や埃等の欠陥による可視データの画素値の低減の程度をRGB毎に表す値である。 The visible defect depth Rf m, n, Gf m, n, Bf m, n , as shown in FIG. 4 (b), R light, G light, and the pixel of the target pixel O in each of the visual data of the B light How much the value is reduced with respect to the average normal pixel values Rave m, n , Gave m, n and Bave m, n that are average values of normal pixels around the target pixel O, that is, scratches, dust, etc. This is a value representing the degree of reduction in the pixel value of visible data due to a defect for each RGB.

そして、以上の処理#03〜#07により演算された対象画素Oの赤外欠陥深度Dfm,n及び可視欠陥深度Rfm,n、Gfm,n、Bfm,nを演算結果記憶部30に記憶する(#08)。 Then, the above process # 03~ # infrared defect depth Df m of the target pixel O calculated by 07, n and visible defects depth Rf m, n, Gf m, n, Bf m, the n operation result storage section 30 (# 08).

その後、欠陥画素マップに登録された全ての欠陥画素が対象画素Oとして選択され、#03〜#08の処理を終了したか否かについて判断する(#09)。欠陥画素マップに登録された全ての欠陥画素について#03〜#08の処理を終了していない場合には(#09:NO)、処理は#03へ戻り、欠陥画素マップに登録された欠陥画素であって未だ対象画素Oとして選択されていない欠陥画素を次の対象画素Oとして選択し(#03)、#04〜#08の処理を上記と同様にして行う。そして、欠陥画素マップに登録された全ての欠陥画素について#03〜#08の処理を終了した場合には(#09:YES)、処理は#10へ進む。この際、演算結果記憶部30には、対象画素Oを順次選択することにより演算された赤外欠陥深度Dfm,n及び可視欠陥深度Rfm,n、Gfm,n、Bfm,nが全て記憶されている。 Thereafter, all the defective pixels registered in the defective pixel map are selected as the target pixel O, and it is determined whether or not the processes of # 03 to # 08 are finished (# 09). If the processes of # 03 to # 08 have not been completed for all the defective pixels registered in the defective pixel map (# 09: NO), the process returns to # 03, and the defective pixels registered in the defective pixel map A defective pixel that has not yet been selected as the target pixel O is selected as the next target pixel O (# 03), and the processes of # 04 to # 08 are performed in the same manner as described above. When the processes of # 03 to # 08 are completed for all defective pixels registered in the defective pixel map (# 09: YES), the process proceeds to # 10. At this time, the operation result storage unit 30, an infrared defect depth Df m calculated by sequentially selecting the target pixel O, n and visible defects depth Rf m, n, Gf m, n, Bf m, n is Everything is remembered.

次に、回帰演算部31において、演算結果記憶部30に記憶された複数の赤外欠陥深度Dfm,nと可視欠陥深度Rfm,n、Gfm,n、Bfm,nとについて回帰分析を行い、回帰係数Ra、Ga、Baと相関係数Rr、Gr、Brとを演算する(#10)。この回帰分析の方法としては、よく知られている最小二乗法を用いた線形回帰分析により好適に行うことができる。本実施形態においては、可視データはR光、G光、及びB光の3色のデータを有しているので、R、G、Bのそれぞれの可視欠陥深度Rfm,n、Gfm,n、Bfm,nについて、赤外欠陥深度Dfm,nとの間で、赤外欠陥深度Dfm,nをX軸、可視欠陥深度Rfm,n、Gfm,n、Bfm,nをY軸として線形回帰分析を行う。この回帰分析の具体例として、R光の可視欠陥深度Rfm,nと赤外欠陥深度Dfm,nとの回帰分析を行い、回帰係数Raと相関係数Rrとを演算する場合の演算式を以下の式(15)〜(19)に示す。 Then, the regression calculation unit 31, the operation result storage unit 30 a plurality of infrared defect depth Df m stored in, n and visible defects depth Rf m, n, Gf m, n, Bf m, regression analysis for the n To calculate regression coefficients Ra, Ga, Ba and correlation coefficients Rr, Gr, Br (# 10). As a method of this regression analysis, linear regression analysis using a well-known least square method can be suitably performed. In the present embodiment, since the visible data includes data of three colors of R light, G light, and B light, the visible defect depths Rfm , n , Gfm , n for R, G, and B, respectively. , Bf m, n , the infrared defect depth Df m, n between the infrared defect depth Df m, n and the X axis, the visible defect depth Rf m, n , Gf m, n , Bf m, n Perform linear regression analysis as Y-axis. As a specific example of this regression analysis, a regression analysis of the visible defect depth Rf m, n of the R light and the infrared defect depth Df m, n is performed, and an equation for calculating the regression coefficient Ra and the correlation coefficient Rr Is shown in the following formulas (15) to (19).

Figure 2005142621
Figure 2005142621

ここで、座標(x,y)は処理対象のNX×NY画素の画像領域上の座標を表す。なお、式(15)〜(19)のΣで表されている和の演算は、NX×NY画素の画像領域上の赤外欠陥深度Dfm,n及び可視欠陥深度Rfm,nが演算された対象画素Oについてのみ行われる。また、kは演算結果記憶部30に記憶された赤外欠陥深度Dfm,n及び可視欠陥深度Rfm,nが演算された対象画素Oの数である。 Here, the coordinates (x, y) represent coordinates on the image area of the NX × NY pixel to be processed. In the calculation of the sum represented by Σ in the equations (15) to (19), the infrared defect depth Df m, n and the visible defect depth Rf m, n on the image area of NX × NY pixels are calculated. This is performed only for the target pixel O. K is the number of target pixels O for which the infrared defect depth Dfm , n and the visible defect depth Rfm , n stored in the computation result storage unit 30 are computed.

G光又はB光の可視欠陥深度Gfm,n、Bfm,nと赤外欠陥深度Dfm,nとの回帰分析を行い、回帰係数Ga、Baと相関係数Gr、Brとを演算する場合の演算式は、以下の式(15)〜(19)におけるRをG又はBに入れ替えたものに等しい。 Perform regression analysis of visible defect depth Gfm , n , Bfm , n and infrared defect depth Dfm , n of G light or B light, and calculate regression coefficients Ga, Ba and correlation coefficients Gr, Br The arithmetic expression in this case is equal to the expression (15) to (19) below in which R is replaced with G or B.

次に、補正係数演算部32において、上記のようにして演算された回帰係数Ra、Ga、Ba及び相関係数Rr、Gr、Brに基づいて可視データと非可視データとの欠陥深度のずれを補正するR光、G光、及びB光の各可視データに共通の補正係数Wを演算する(#11)。具体的には、以下の式(20)に従ってR光、G光、及びB光の各色の相関係数Rr、Gr、Brを当該色の回帰係数Ra、Ga、Baの重みとするRGBの重み付き平均値を演算して共通回帰係数w1するとともに、式(21)に従って各色の相関係数Rr、Gr、Brの単純平均値を演算して共通相関係数w2とし、式(22)に従って共通回帰係数w1と所定の基準値γとの差に対して共通相関係数w2を乗算して得られた値に基準値γを加算して補正係数Wを演算する。   Next, in the correction coefficient calculation unit 32, the deviation of the defect depth between the visible data and the invisible data is calculated based on the regression coefficients Ra, Ga, Ba and the correlation coefficients Rr, Gr, Br calculated as described above. A correction coefficient W common to the visible data of R light, G light, and B light to be corrected is calculated (# 11). Specifically, according to the following equation (20), RGB weights using the correlation coefficients Rr, Gr, Br of the colors of R light, G light, and B light as the weights of the regression coefficients Ra, Ga, Ba of the color. A common regression coefficient w1 is calculated by calculating the attached average value, and a simple average value of the correlation coefficients Rr, Gr, and Br of each color is calculated as the common correlation coefficient w2 according to the equation (21), and is common according to the equation (22). The correction coefficient W is calculated by adding the reference value γ to the value obtained by multiplying the difference between the regression coefficient w1 and the predetermined reference value γ by the common correlation coefficient w2.

Figure 2005142621
Figure 2005142621

ここで、経験上、可視データの方が赤外データに比べて傷や埃等の欠陥に対して敏感に反応し、同じ欠陥に対する欠陥深度は、可視欠陥深度Rfm,n、Gfm,n、Bfm,nの方が赤外欠陥深度Dfm,nよりも大きくなることが分かっている。したがって、赤外欠陥深度Dfm,nをX軸、可視欠陥深度Rfm,n、Gfm,n、Bfm,nをY軸として線形回帰分析を行う場合における回帰直線の傾きに相当する回帰係数Ra、Ga、Baは通常1以上となる。そこで、回帰係数Ra、Ga、Baが1未満となった場合には、その演算結果は誤りである可能性が高いことから回帰係数Ra、Ga、Baの値を1に固定することとし、その際には相関係数Rr、Gr、Brの値も1として上記式(20)及び(21)の演算を行う。また、本実施形態においては、補正係数Wによる可視データと非可視データとの欠陥深度のずれの補正量がゼロとなる補正係数Wの値すなわち「1」を基準値γの値としている。そして、共通回帰係数w1の基準値γ(=1)を超える部分に対して共通相関係数w2を乗算することにより、補正係数Wによる前記補正量に相当する部分(w1−γ)に対して当該補正量の正確性に相当する値(w2)を乗算することになり、共通相関係数w2の値が高く共通回帰係数w1の正確性が高い場合には共通回帰係数w1が補正係数Wに与える影響を大きくし、逆に共通相関係数w2の値が低く共通回帰係数w1の正確性が低い場合には共通回帰係数w1が補正係数Wに与える影響を小さくすることができる。 Here, empirically, visible data reacts more sensitively to defects such as scratches and dust than infrared data, and the defect depth for the same defect is visible defect depth Rf m, n , Gf m, n , Bf m, towards the n is found to be greater infrared defect depth Df m, than n. Therefore, X-axis infrared defect depth Df m, n, the visible defect depth Rf m, n, Gf m, n, Bf m, n and corresponds to the slope of the regression line in a case of performing linear regression analysis as the Y-axis regression The coefficients Ra, Ga and Ba are usually 1 or more. Therefore, when the regression coefficients Ra, Ga, Ba are less than 1, the calculation result is likely to be incorrect, so the values of the regression coefficients Ra, Ga, Ba are fixed to 1. In this case, the values of the correlation coefficients Rr, Gr, and Br are also set to 1, and the above expressions (20) and (21) are calculated. Further, in the present embodiment, the value of the correction coefficient W at which the correction amount of the defect depth shift between the visible data and the invisible data by the correction coefficient W becomes zero, that is, “1” is set as the reference value γ. Then, by multiplying the portion of the common regression coefficient w1 exceeding the reference value γ (= 1) by the common correlation coefficient w2, the portion corresponding to the correction amount by the correction coefficient W (w1−γ) is obtained. The value (w2) corresponding to the accuracy of the correction amount will be multiplied. When the value of the common correlation coefficient w2 is high and the accuracy of the common regression coefficient w1 is high, the common regression coefficient w1 is added to the correction coefficient W. On the contrary, when the value of the common correlation coefficient w2 is low and the accuracy of the common regression coefficient w1 is low, the influence of the common regression coefficient w1 on the correction coefficient W can be reduced.

なお、補正係数演算部32における補正係数の演算方法はこれに限定されるものではなく、例えば、若干補正の度合いが強くなりすぎる場合もあるが、式(20)の結果をそのまま補正係数Wとして用いることも可能である。   The correction coefficient calculation method in the correction coefficient calculation unit 32 is not limited to this. For example, the degree of correction may be slightly too strong, but the result of equation (20) is used as the correction coefficient W as it is. It is also possible to use it.

その後、修正部33において、補正係数Wと各欠陥画素の赤外欠陥深度Dfm,nとに基づいて各欠陥画素の輝度調整による修正を行う(#12)。ここでは、欠陥画素マップに登録された複数の欠陥画素を順次選択して輝度調整による修正を行うこととし、当該選択された欠陥画素の座標は上記対象画素Oと同様に(m,n)とする。この輝度調整による修正の処理は、具体的には、以下の式(26)〜(28)に従って、R光、G光、及びB光の各可視データ毎に、欠陥画素の画素値Rdatm,n、Gdatm,n、Bdatm,nに対して補正係数Wにより補正した赤外欠陥深度Dfm,nを加算することにより行う。 Thereafter, the correction unit 33 performs correction by adjusting the luminance of each defective pixel based on the correction coefficient W and the infrared defect depth Dfm , n of each defective pixel (# 12). Here, a plurality of defective pixels registered in the defective pixel map are sequentially selected and corrected by luminance adjustment, and the coordinates of the selected defective pixels are (m, n) as in the target pixel O. To do. Specifically, the correction processing by brightness adjustment is performed according to the following formulas (26) to (28) for each visible data of R light, G light, and B light, the pixel value Rdat m, This is performed by adding the infrared defect depth Df m, n corrected by the correction coefficient W to n 1 , Gdat m, n and Bdat m, n .

Figure 2005142621
Figure 2005142621

ここで、赤外欠陥深度Dfm,nに対して補正係数Wを乗算することにより補正した値(Dfm,n×W)は、可視光と非可視光との波長或いは屈折率の相違に起因する、可視データが欠陥により受ける影響と非可視データが欠陥により受ける影響とのずれを考慮して、赤外欠陥深度Dfm,nを、可視データの欠陥深度に合せて補正した値に相当する。したがって、この赤外欠陥深度Dfm,nに対して補正係数Wを乗算することにより補正した値(Dfm,n×W)を欠陥画素の画素値Rdatm,n、Gdatm,n、Bdatm,nに加算することにより、当該欠陥画素の輝度を上げて修正することができる。 Here, the value (Df m, n × W) corrected by multiplying the infrared defect depth Df m, n by the correction coefficient W is the difference in wavelength or refractive index between visible light and invisible light. Considering the difference between the effect of visible data caused by defects and the effect of non-visible data caused by defects, the infrared defect depth Df m, n is equivalent to the value corrected according to the defect depth of visible data To do. Accordingly, the value (Df m, n × W) corrected by multiplying the infrared defect depth Df m, n by the correction coefficient W is used as the pixel value Rdat m, n , Gdat m, n , Bdat of the defective pixel. By adding to m, n , the brightness of the defective pixel can be increased and corrected.

なお、上記実施形態においては、補正係数演算部32における補正係数Wの演算について、上記の式(20)〜(22)に示すように、傷や埃等の欠陥による欠陥深度のずれは、可視光(RGB)と赤外光とのずれにくらべて可視光同士のずれは非常に小さいことに鑑み、以降の処理#12における輝度調整による修正処理の演算量を低減するために、R光、G光、及びB光の各可視データに共通の一の補正係数Wを演算している。しかし、より正確に欠陥深度のずれを補正するためには、可視光(RGB)と赤外光とのずれだけでなく、可視光同士のずれも考慮した補正を行うことが必要となる場合もある。そこで、R光、G光、及びB光の各可視データ毎に補正係数Wr、Wg、Wbをそれぞれ演算することも可能である。この場合、RGBの各色毎に回帰係数Ra、Ga、Baと所定の基準値γとの差に対して当該色の相関係数Rr、Gr、Brを乗算して得られた値に基準値γを加算して各色の補正係数Wr、Wg、Wbをそれぞれ演算する。具体的には、以下の式(23)〜(25)に従って演算する。   In the above embodiment, regarding the calculation of the correction coefficient W in the correction coefficient calculation unit 32, as shown in the above formulas (20) to (22), the deviation of the defect depth due to defects such as scratches and dust is visible. In view of the fact that the shift between visible lights is very small compared with the shift between light (RGB) and infrared light, in order to reduce the amount of correction processing by brightness adjustment in the subsequent process # 12, R light, One correction coefficient W common to the visible data of G light and B light is calculated. However, in order to correct the deviation of the defect depth more accurately, it may be necessary to perform correction in consideration of not only the deviation between visible light (RGB) and infrared light but also the deviation between visible lights. is there. Therefore, it is also possible to calculate correction coefficients Wr, Wg, and Wb for each visible data of R light, G light, and B light. In this case, the value obtained by multiplying the difference between the regression coefficients Ra, Ga, Ba and the predetermined reference value γ for each color of RGB by the correlation coefficient Rr, Gr, Br of the color is the reference value γ. Are added to calculate correction coefficients Wr, Wg, and Wb for each color. Specifically, the calculation is performed according to the following equations (23) to (25).

Figure 2005142621
Figure 2005142621

また、上記実施形態においては、可視データがR光、G光、及びB光の3色を有するカラー画像を扱う場合について説明したが、本発明は、可視データが単色の場合、すなわちモノクロ画像の場合にも同様に適用することが可能である。その場合、平均正常画素値演算部28において各正常画素の重みを演算する際には、修正正常画素値のRGB平均値及び修正対象画素値のRGB平均値に代えて、当該単色の一の可視データについての修正正常画素値及び修正対象画素値を用いることになる。また、補正係数演算部32において補正係数を演算する際には、当該単色の一の可視データについての回帰係数と所定の基準値γとの差に対して当該単色の一の可視データについての相関係数を乗算して得られた値に基準値γを加算して補正係数Wを演算することになる。   In the above embodiment, the case where the visible data handles a color image having three colors of R light, G light, and B light has been described. However, the present invention provides a case where the visible data is a single color, that is, a monochrome image. The same can be applied to the case. In this case, when calculating the weight of each normal pixel in the average normal pixel value calculation unit 28, instead of the RGB average value of the corrected normal pixel value and the RGB average value of the correction target pixel value, one visible color of the single color is displayed. The corrected normal pixel value and the correction target pixel value for the data are used. In addition, when calculating the correction coefficient in the correction coefficient calculation unit 32, the phase of the visible data of the single color with respect to the difference between the regression coefficient of the visible data of the single color and the predetermined reference value γ. The correction coefficient W is calculated by adding the reference value γ to the value obtained by multiplying the number of relations.

本発明は、写真フィルム等の画像記録材料に記録した画像をデジタル画像データとして読み取った後、印画紙に記録する画像プリントシステム等の画像処理装置、方法、又はプログラムとして好適に利用することができる。   INDUSTRIAL APPLICABILITY The present invention can be suitably used as an image processing apparatus, method, or program such as an image print system that reads an image recorded on an image recording material such as a photographic film as digital image data and then records the image on a photographic paper. .

本実施形態に係る画像プリントシステムの外観を示す斜視図The perspective view which shows the external appearance of the image printing system which concerns on this embodiment 本実施形態に係る画像プリントシステムの概略構成を示す模式図Schematic diagram showing a schematic configuration of an image print system according to the present embodiment 本実施形態に係る画像処理装置の機能を示すブロック図The block diagram which shows the function of the image processing apparatus which concerns on this embodiment 画像データ中の一つの直線Lに沿って配列された画素の画素値を、赤外データ及びR光、G光、及びB光のそれぞれの可視データについて表したグラフA graph representing pixel values of pixels arranged along one straight line L in image data with respect to infrared data and visible data of R light, G light, and B light, respectively. 本実施形態に係る画像処理装置における欠陥画素の修正処理のフローチャートFlowchart of defective pixel correction processing in the image processing apparatus according to the present embodiment 対象画素とその周囲の一定領域の一例を示す説明図Explanatory drawing showing an example of the target pixel and its surrounding area 本実施形態に係る画像処理装置における平均正常画素値の演算処理を示すフローチャートThe flowchart which shows the calculation process of the average normal pixel value in the image processing apparatus which concerns on this embodiment.

符号の説明Explanation of symbols

1 画像処理装置
2 写真フィルム
4 画像プリントシステム
5 画像読取装置
6 画像記録装置
19 欠陥修正部
18 メモリ
25 欠陥画素検出部
26 対象画素選択部
27 赤外欠陥深度演算部
28 平均正常画素値演算部
29 可視欠陥深度演算部
30 演算結果記憶部
31 回帰演算部
32 補正係数演算部
33 修正部
O 対象画素
E 対象画素の周囲の一定領域
CF 赤外データの基準画素値
Dfm,n 赤外欠陥深度
Ravem,n、Gavem,n、Bavem,n 平均正常画素値
Rfm,n、Gfm,n、Bfm,n 可視欠陥深度
Ra、Ga、Ba 回帰係数
Rr、Gr、Br 相関係数
W 補正係数
DESCRIPTION OF SYMBOLS 1 Image processing apparatus 2 Photo film 4 Image printing system 5 Image reader 6 Image recording apparatus 19 Defect correction part 18 Memory 25 Defect pixel detection part 26 Target pixel selection part 27 Infrared defect depth calculation part 28 Average normal pixel value calculation part 29 Visible defect depth calculation unit 30 Calculation result storage unit 31 Regression calculation unit 32 Correction coefficient calculation unit 33 Correction unit O Target pixel E Constant region around the target pixel
Reference pixel value of CF infrared data
Df m, n infrared defect depth
Rave m, n , Gave m, n , Bave m, n Average normal pixel value
Rf m, n , Gf m, n , Bf m, n Visible defect depth
Ra, Ga, Ba regression coefficients
Rr, Gr, Br correlation coefficient
W correction factor

Claims (7)

画像記録材料に対して可視光及び非可視光を照射して可視データ及び非可視データをそれぞれ取得し、前記非可視データに基づいて検出された前記可視データに含まれる欠陥画素の修正を行う画像処理装置において、
複数の前記欠陥画素の一つを対象画素として選択する対象画素選択手段と、
前記非可視データの基準画素値と前記非可視データの前記対象画素に対応する画素の画素値とに基づいて前記対象画素の非可視データに基づく欠陥深度を演算する非可視欠陥深度演算手段と、
前記可視データにおける前記対象画素の周囲の一定領域内の正常画素の平均画素値を演算する平均正常画素値演算手段と、
前記対象画素の画素値と前記平均正常画素値とに基づいて前記対象画素の可視データに基づく欠陥深度を演算する可視欠陥深度演算手段と、
前記対象画素を順次選択することにより前記非可視欠陥深度演算手段及び可視欠陥深度演算手段においてそれぞれ複数演算される前記非可視欠陥深度及び可視欠陥深度を記憶する記憶手段と、
前記記憶手段に記憶された複数の前記非可視欠陥深度と前記可視欠陥深度とについて回帰分析を行い、回帰係数と相関係数とを演算する回帰演算手段と、
前記回帰係数及び相関係数に基づいて前記可視データと非可視データとの欠陥深度のずれを補正する補正係数を演算する補正係数演算手段と、
前記補正係数と各欠陥画素の非可視欠陥深度とに基づいて各欠陥画素の輝度調整による修正を行う修正手段とを有する画像処理装置。
An image for irradiating an image recording material with visible light and non-visible light to obtain visible data and non-visible data, respectively, and correcting defective pixels included in the visible data detected based on the non-visible data In the processing device,
Target pixel selection means for selecting one of the plurality of defective pixels as a target pixel;
Non-visible defect depth calculation means for calculating a defect depth based on non-visible data of the target pixel based on a reference pixel value of the non-visible data and a pixel value of a pixel corresponding to the target pixel of the non-visible data;
An average normal pixel value calculating means for calculating an average pixel value of normal pixels in a fixed region around the target pixel in the visible data;
Visible defect depth calculation means for calculating a defect depth based on visible data of the target pixel based on the pixel value of the target pixel and the average normal pixel value;
Storage means for storing the invisible defect depth and the visible defect depth, which are respectively calculated in the invisible defect depth calculating means and the visible defect depth calculating means by sequentially selecting the target pixels;
Regression calculation means for performing regression analysis on the plurality of invisible defect depths and the visible defect depths stored in the storage means, and calculating regression coefficients and correlation coefficients;
Correction coefficient calculation means for calculating a correction coefficient for correcting a shift in defect depth between the visible data and the invisible data based on the regression coefficient and the correlation coefficient;
An image processing apparatus comprising: a correcting unit that corrects each defective pixel by brightness adjustment based on the correction coefficient and the invisible defect depth of each defective pixel.
前記補正係数演算手段は、前記回帰係数と所定の基準値との差に対して相関係数を乗算して得られた値に前記基準値を加算して補正係数を演算する請求項1に記載の画像処理装置。   The correction coefficient calculation means calculates the correction coefficient by adding the reference value to a value obtained by multiplying a difference between the regression coefficient and a predetermined reference value by a correlation coefficient. Image processing apparatus. 前記補正係数演算手段は、複数色の可視データに共通する前記補正係数を演算する場合に、各色の前記相関係数を当該色の前記回帰係数の重みとする複数色の重み付き平均値を演算して共通回帰係数とし、複数色の前記相関係数の平均値を演算して共通相関係数とし、これらの共通回帰係数及び共通相関係数を、前記回帰係数及び相関係数に代えて用いて補正係数の演算を行う請求項1又は2に記載の画像処理装置。   When calculating the correction coefficient common to visible data of a plurality of colors, the correction coefficient calculation means calculates a weighted average value of a plurality of colors using the correlation coefficient of each color as the weight of the regression coefficient of the color The common regression coefficient is calculated, and an average value of the correlation coefficients of a plurality of colors is calculated to be a common correlation coefficient. These common regression coefficient and common correlation coefficient are used instead of the regression coefficient and the correlation coefficient. The image processing apparatus according to claim 1, wherein the correction coefficient is calculated. 前記平均正常画素値演算手段は、前記非可視データの基準画素値と前記非可視データの前記各正常画素に対応する画素の画素値とに基づいて前記一定領域内の各正常画素の非可視データに基づく深度修正値を演算し、この各正常画素の深度修正値により当該各正常画素の画素値を輝度調整して演算される修正正常画素値と、前記対象画素の非可視欠陥深度により前記対象画素の画素値を輝度調整して演算される修正対象画素値との差の絶対値の所定値に対する補数を演算して各正常画素の重みとし、前記一定領域内の正常画素の修正正常画素値の前記重みを用いた重み付き平均値を演算し、その演算結果を平均正常画素値とする請求項1から3の何れか1項に記載の画像処理装置。   The average normal pixel value calculation means is configured to determine the invisible data of each normal pixel in the certain area based on a reference pixel value of the invisible data and a pixel value of a pixel corresponding to each normal pixel of the invisible data. The corrected normal pixel value calculated by calculating the brightness correction value based on the pixel value and adjusting the luminance of the pixel value of each normal pixel by the depth correction value of each normal pixel and the invisible defect depth of the target pixel The corrected normal pixel value of the normal pixel in the certain area is calculated by calculating the complement of the absolute value of the difference from the correction target pixel value calculated by adjusting the luminance of the pixel value of the pixel as a weight of each normal pixel. 4. The image processing apparatus according to claim 1, wherein a weighted average value using the weights is calculated, and the calculation result is an average normal pixel value. 5. 前記平均正常画素値演算手段は、複数色の可視データに基づいて各正常画素の重みを演算する際に、前記修正正常画素値に代えて各色の修正正常画素値の平均値を用い、前記修正対象画素値に代えて各色の修正対象画素値の平均値を用いる請求項4に記載の画像処理装置。   The average normal pixel value calculation means uses the average value of the corrected normal pixel value of each color instead of the corrected normal pixel value when calculating the weight of each normal pixel based on visible data of a plurality of colors, and the correction The image processing apparatus according to claim 4, wherein an average value of correction target pixel values of each color is used instead of the target pixel value. 画像記録材料に対して可視光及び非可視光を照射して可視データ及び非可視データをそれぞれ取得し、前記非可視データに基づいて検出された前記可視データに含まれる欠陥画素の修正を行う画像処理方法において、
複数の前記欠陥画素の一つを対象画素として選択し、前記非可視データの基準画素値と前記非可視データの前記対象画素に対応する画素の画素値とに基づいて前記対象画素の非可視データに基づく欠陥深度を演算し、前記可視データにおける前記対象画素の周囲の一定領域内の正常画素の平均画素値を演算し、前記対象画素の画素値と前記平均正常画素値とに基づいて前記対象画素の可視データに基づく欠陥深度を演算し、前記対象画素を順次選択することによりそれぞれ複数演算される前記非可視欠陥深度及び可視欠陥深度について回帰分析を行って回帰係数と相関係数とを演算し、前記回帰係数及び相関係数に基づいて前記可視データと非可視データとの欠陥深度のずれを補正する補正係数を演算し、前記補正係数と各欠陥画素の非可視欠陥深度とに基づいて各欠陥画素の輝度調整による修正を行う画像処理方法。
An image for irradiating an image recording material with visible light and non-visible light to obtain visible data and non-visible data, respectively, and correcting defective pixels included in the visible data detected based on the non-visible data In the processing method,
One of the plurality of defective pixels is selected as a target pixel, and the non-visible data of the target pixel is based on a reference pixel value of the non-visible data and a pixel value of a pixel corresponding to the target pixel of the non-visible data The defect depth based on the target pixel is calculated, the average pixel value of normal pixels in a certain area around the target pixel in the visible data is calculated, and the target based on the pixel value of the target pixel and the average normal pixel value Calculates the regression coefficient and correlation coefficient by calculating the defect depth based on the visible data of the pixels and performing regression analysis on the invisible defect depth and the visible defect depth, each of which is calculated by sequentially selecting the target pixel. And calculating a correction coefficient for correcting a shift in the defect depth between the visible data and the invisible data based on the regression coefficient and the correlation coefficient, so that the correction coefficient and the invisible of each defective pixel are calculated. An image processing method for performing correction by the luminance adjustment of each defective pixel based on the Recessed depth.
画像記録材料に対して可視光及び非可視光を照射して可視データ及び非可視データをそれぞれ取得し、前記非可視データに基づいて検出された前記可視データに含まれる欠陥画素の修正を行う画像処理方法において、
複数の前記欠陥画素の一つを対象画素として選択し、前記非可視データの基準画素値と前記非可視データの前記対象画素に対応する画素の画素値とに基づいて前記対象画素の非可視データに基づく欠陥深度を演算し、前記可視データにおける前記対象画素の周囲の一定領域内の正常画素の平均画素値を演算し、前記対象画素の画素値と前記平均正常画素値とに基づいて前記対象画素の可視データに基づく欠陥深度を演算し、前記対象画素を順次選択することによりそれぞれ複数演算される前記非可視欠陥深度及び可視欠陥深度について回帰分析を行って回帰係数と相関係数とを演算し、前記回帰係数及び相関係数に基づいて前記可視データと非可視データとの欠陥深度のずれを補正する補正係数を演算し、前記補正係数と各欠陥画素の非可視欠陥深度とに基づいて各欠陥画素の輝度調整による修正を行う処理をコンピュータに実行させるための画像処理プログラム。
An image for irradiating an image recording material with visible light and non-visible light to obtain visible data and non-visible data, respectively, and correcting defective pixels included in the visible data detected based on the non-visible data In the processing method,
One of the plurality of defective pixels is selected as a target pixel, and the non-visible data of the target pixel is based on a reference pixel value of the non-visible data and a pixel value of a pixel corresponding to the target pixel of the non-visible data The defect depth based on the target pixel is calculated, the average pixel value of normal pixels in a certain area around the target pixel in the visible data is calculated, and the target based on the pixel value of the target pixel and the average normal pixel value Calculates the regression coefficient and correlation coefficient by calculating the defect depth based on the visible data of the pixels and performing regression analysis on the invisible defect depth and the visible defect depth, each of which is calculated by sequentially selecting the target pixel. And calculating a correction coefficient for correcting a shift in the defect depth between the visible data and the invisible data based on the regression coefficient and the correlation coefficient, so that the correction coefficient and the invisible of each defective pixel are calculated. The image processing program for executing Recessed processing for performing correction by the luminance adjustment of each defective pixel based on the depth computer.
JP2003374097A 2003-11-04 2003-11-04 Image processing apparatus, method, and program Expired - Lifetime JP3997486B2 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
JP2003374097A JP3997486B2 (en) 2003-11-04 2003-11-04 Image processing apparatus, method, and program

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
JP2003374097A JP3997486B2 (en) 2003-11-04 2003-11-04 Image processing apparatus, method, and program

Publications (2)

Publication Number Publication Date
JP2005142621A true JP2005142621A (en) 2005-06-02
JP3997486B2 JP3997486B2 (en) 2007-10-24

Family

ID=34685914

Family Applications (1)

Application Number Title Priority Date Filing Date
JP2003374097A Expired - Lifetime JP3997486B2 (en) 2003-11-04 2003-11-04 Image processing apparatus, method, and program

Country Status (1)

Country Link
JP (1) JP3997486B2 (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9202266B2 (en) 2012-12-31 2015-12-01 Nokia Technologies Oy Method, apparatus and computer program product for processing of images
CN116074495A (en) * 2023-03-07 2023-05-05 合肥埃科光电科技股份有限公司 Storage method, detection and correction method and device for dead pixel of image sensor

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9202266B2 (en) 2012-12-31 2015-12-01 Nokia Technologies Oy Method, apparatus and computer program product for processing of images
CN116074495A (en) * 2023-03-07 2023-05-05 合肥埃科光电科技股份有限公司 Storage method, detection and correction method and device for dead pixel of image sensor

Also Published As

Publication number Publication date
JP3997486B2 (en) 2007-10-24

Similar Documents

Publication Publication Date Title
JP3952301B2 (en) Image processing apparatus, method, and program
JP3938120B2 (en) Image processing apparatus, method, and program
JP4244018B2 (en) Defective pixel correction method, program, and defective pixel correction system for implementing the method
US7433545B2 (en) Image processing apparatus and image processing method for correcting image data
US7356167B2 (en) Image processing apparatus and image processing method for correcting image data
JP3997486B2 (en) Image processing apparatus, method, and program
JP2006324987A (en) Photographic image processing method, photographic image processing program, and photographic image processing module
EP1562144B1 (en) Image processing apparatus and image processing method for correcting image data
JP2005063022A (en) Noise pixel map producing method, device performing the same, program, and photograph printing device
JP4400481B2 (en) Image processing device
JP4284604B2 (en) Contrast adjustment method and contrast adjustment apparatus for implementing the method
JP4513006B2 (en) Emulsion scratch determination method and emulsion scratch determination system
JP2005078314A (en) Position adjusting method and image reading device with stain compensation function using the same
JP4586457B2 (en) Image processing device
JP2007235737A (en) Defective pixel correction method, and defective pixel correction system
JP4560720B2 (en) Emulsion scratch determination method and emulsion scratch determination system
JP2005091434A (en) Position adjusting method and image reader with damage compensation function using the same
JP2006024013A (en) Film information code decoding method and system for implementing the method
JP2006186929A (en) Image processing apparatus
JP2006080616A (en) Transportation fluctuation measuring apparatus and measuring program
JP2006174349A (en) Method and processor for image repair

Legal Events

Date Code Title Description
A621 Written request for application examination

Free format text: JAPANESE INTERMEDIATE CODE: A621

Effective date: 20060912

A131 Notification of reasons for refusal

Free format text: JAPANESE INTERMEDIATE CODE: A131

Effective date: 20061221

A521 Request for written amendment filed

Free format text: JAPANESE INTERMEDIATE CODE: A523

Effective date: 20070219

TRDD Decision of grant or rejection written
A01 Written decision to grant a patent or to grant a registration (utility model)

Free format text: JAPANESE INTERMEDIATE CODE: A01

Effective date: 20070712

A61 First payment of annual fees (during grant procedure)

Free format text: JAPANESE INTERMEDIATE CODE: A61

Effective date: 20070725

R150 Certificate of patent or registration of utility model

Free format text: JAPANESE INTERMEDIATE CODE: R150

FPAY Renewal fee payment (event date is renewal date of database)

Free format text: PAYMENT UNTIL: 20100817

Year of fee payment: 3

FPAY Renewal fee payment (event date is renewal date of database)

Free format text: PAYMENT UNTIL: 20110817

Year of fee payment: 4

FPAY Renewal fee payment (event date is renewal date of database)

Free format text: PAYMENT UNTIL: 20110817

Year of fee payment: 4

S801 Written request for registration of abandonment of right

Free format text: JAPANESE INTERMEDIATE CODE: R311801

ABAN Cancellation due to abandonment
FPAY Renewal fee payment (event date is renewal date of database)

Free format text: PAYMENT UNTIL: 20110817

Year of fee payment: 4

R350 Written notification of registration of transfer

Free format text: JAPANESE INTERMEDIATE CODE: R350