CN1806449A - Image processing apparatus, image processing method, and computer program - Google Patents
Image processing apparatus, image processing method, and computer program Download PDFInfo
- Publication number
- CN1806449A CN1806449A CN200580000559.3A CN200580000559A CN1806449A CN 1806449 A CN1806449 A CN 1806449A CN 200580000559 A CN200580000559 A CN 200580000559A CN 1806449 A CN1806449 A CN 1806449A
- Authority
- CN
- China
- Prior art keywords
- pixel
- image
- colour element
- false
- false colour
- 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
Links
- 238000012545 processing Methods 0.000 title claims description 244
- 238000003672 processing method Methods 0.000 title claims description 35
- 238000004590 computer program Methods 0.000 title claims description 10
- 241001326510 Phacelia sericea Species 0.000 claims abstract description 168
- 238000001514 detection method Methods 0.000 claims abstract description 37
- 229920006395 saturated elastomer Polymers 0.000 claims description 252
- 238000006243 chemical reaction Methods 0.000 claims description 38
- 230000035945 sensitivity Effects 0.000 claims description 19
- 238000004458 analytical method Methods 0.000 claims description 8
- 238000013519 translation Methods 0.000 claims description 5
- 238000000034 method Methods 0.000 abstract description 51
- 238000012937 correction Methods 0.000 abstract description 14
- 230000004075 alteration Effects 0.000 abstract description 13
- 239000003086 colorant Substances 0.000 abstract description 5
- 230000000875 corresponding effect Effects 0.000 description 70
- 230000008569 process Effects 0.000 description 40
- 230000003287 optical effect Effects 0.000 description 26
- 238000001914 filtration Methods 0.000 description 14
- 238000010586 diagram Methods 0.000 description 13
- 230000008859 change Effects 0.000 description 11
- 238000012958 reprocessing Methods 0.000 description 7
- 230000001915 proofreading effect Effects 0.000 description 5
- 230000009467 reduction Effects 0.000 description 5
- 239000000284 extract Substances 0.000 description 4
- 238000005096 rolling process Methods 0.000 description 3
- 241001269238 Data Species 0.000 description 2
- 230000002596 correlated effect Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 238000005070 sampling Methods 0.000 description 2
- 239000004065 semiconductor Substances 0.000 description 2
- 206010010071 Coma Diseases 0.000 description 1
- 240000004859 Gamochaeta purpurea Species 0.000 description 1
- 230000009471 action Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- WUKWITHWXAAZEY-UHFFFAOYSA-L calcium difluoride Chemical compound [F-].[F-].[Ca+2] WUKWITHWXAAZEY-UHFFFAOYSA-L 0.000 description 1
- 235000019646 color tone Nutrition 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 230000001447 compensatory effect Effects 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 230000006866 deterioration Effects 0.000 description 1
- 235000013399 edible fruits Nutrition 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 239000012467 final product Substances 0.000 description 1
- 239000010436 fluorite Substances 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 239000011521 glass Substances 0.000 description 1
- 230000010365 information processing Effects 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 230000014759 maintenance of location Effects 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000005693 optoelectronics Effects 0.000 description 1
- 230000002093 peripheral effect Effects 0.000 description 1
- 238000012163 sequencing technique Methods 0.000 description 1
Images
Landscapes
- Facsimile Image Signal Circuits (AREA)
- Image Processing (AREA)
Abstract
An apparatus and method for efficiently executing a correction of false colors, such as purple fringe or the like, caused by chromatic aberration to produce and output high quality image data. White non-gradation pixels are detected from image data, and a false color pixel detected area is established around the detected white non-gradation pixels. In the established area, pixels having a color corresponding to the false color such as purple fringe or the like are detected, and the detected pixels are specified as false color pixels. The specified false color pixels are corrected based on a periphery pixel value. This arrangement allows an efficient detection of false color area of the purple fringe or the like occurring in the vicinity of the white non-gradation pixels, and also allows a partial pixel value correction, thereby allowing the production and output of high quality image data without affecting the whole image.
Description
Technical field
The present invention relates to image processing equipment, image processing method and computer program.Particularly, the present invention relates to be used for by proofreading and correct wherein for example because lens aberration produces image processing equipment, image processing method and the computer program that the view data of false colored (false color) produces the high quality graphic data.
Background technology
For using video camera to take, owing to lens aberration produces variety of issue.The example of typical chromatic aberation comprises 5 aberrations of Saden you (Seidel), and they are monochromatic aberrations.This is that you analyze and based on 5 aberrations of lens sphere by the Saden of Germany, and is the generic term of spherical aberration, coma aberration, astigmatic image error, distorton aberration and field curvature.Except those aberrations, known chromatic aberation can also cause serious problem.Chromatic aberation results from the fact that refractive index with respect to lens material light changes according to its wavelength, and described chromatic aberation produces false colored on image capture surface.
The typical case of chromatic aberation, axial chromatic aberation and expansive color aberration is known, because the focal position on optical axis changes according to wavelength, therefore described axial chromatic aberation causes that color blurs (color blur), and change according to wavelength because image amplifies, so the expansive color aberration causes gamut (color shift).Generally, the phenomenon that is called as " purple fringe (purple fringe) " in English also is the major reason of image deterioration.This is such phenomenon: wherein, because the difference in the point that causes owing to the optical wavelength expansion (pointspread) therefore produces false colored in the marginal portion in image.Even when it is not visible in typical pixel, the existence that the high-contrast edges part of saturated in vain (white saturation) takes place also makes and produces the false colour of purple at described marginal portion periphery, therefore form factitious image, the saturated state of described white saturated i.e. wherein brightness (luminance) level.Generally, the false colour of generation is called as " purple fringe " near white saturated part, because produce a large amount of colors based on purple.But according to lens and shooting condition, described false colour can be any color, such as having green color.Below, " purple fringe " refers to the false colour that partly occurs in the white saturated high-contrast edges of generation and leads to a phenomenon, and irrelevant with the color that is produced.
As the technology that is used to reduce chromatic aberation, use the lens of the glass that constitutes with the special material such as fluorite to use.But because need expensively make such lens, so price is also high.Such lens are used for some the senior video cameras such as the video camera of replaceable lens, but can not be widely used.
But in the processing described in the patent document 1, because pass through to reduce the saturated vacation colour that reduces of color, the color that has therefore also reduced the primary object color is saturated, the feasible thus original color that can not reproduce object faithfully.As a result, exist output to seem the problem of factitious image.In addition, in order to estimate the brightness of white saturated part, need take for twice.When taking place during twice shooting operation that hand moves or object when moving, there is the problem that is difficult to obtain correct result.
[patent document 1] Japanese unexamined patent discloses 2003-60983 number.
Summary of the invention
Make the present invention based on the problems referred to above, the purpose of this invention is to provide a kind of image processing equipment, image processing method and computer program, they are colored by the vacation of paying close attention to such as the purple fringe that produces at white saturated part periphery, detect false colored region effectively, and operating part is proofreaied and correct, and feasible can the generation with the outputting high quality view data do not influence entire image.
A first aspect of the present invention provides a kind of image processing equipment.
Described image processing equipment comprises:
White saturation detection unit is used for detecting white saturated pixel from view data;
False colour element surveyed area is provided with the unit, is used for being provided with around the white saturated pixel that is detected by white saturation detection unit false colour element surveyed area;
False colored detecting unit is used for the zone that the unit is provided with is being set by false colour element surveyed area, and the pixel that will have the color corresponding with false colour is identified as false colour element; And
The pixel value correcting unit is used for the value according to surrounding pixel, and the false colour element that is detected by the colored detecting unit of vacation is carried out the pixel value treatment for correcting.
In addition, according to an embodiment of image processing equipment of the present invention, the pixel value correcting unit has: the compensated interpolation unit is used for false colour element being carried out compensated interpolation handling according to pixel value on every side; And color Fuzzy Processing unit, be used for false colour element is carried out the color Fuzzy Processing.
In addition, an embodiment according to image processing equipment of the present invention, the pixel value correcting unit is selected the pixel except false colour element and white saturated pixel, and is carried out compensated interpolation according to selected pixel and handle from the pixel that exists around false colour element.
In addition, according to an embodiment of image processing equipment of the present invention, false colored detecting unit is carried out the processing that the pixel that will have default specific color in the zone that the unit setting is set by false colour element surveyed area is identified as false colour element.
In addition, an embodiment according to image processing equipment of the present invention, described false colour is a purple fringe, and false colored detecting unit is carried out the processing that the purple pixel in the zone that the unit setting is set by false colour element surveyed area is identified as false colour element.
In addition, an embodiment according to image processing equipment of the present invention, described false colour element surveyed area is provided with the unit, according to take with the aperture during the processed view data, focus information and from light center to white saturated pixel (x, among the data of distance y) at least one carried out the processing of determining the false colour element surveyed area of setting around the white saturated pixel that is detected by white saturation detection unit.
In addition, according to an embodiment of image processing equipment of the present invention, the processing that selection has the white saturated pixel of the brightness that is greater than or equal to predetermined threshold is carried out in described white saturation detection unit.
In addition, an embodiment according to image processing equipment of the present invention, image processing equipment also comprises the color conversion cell that is used to carry out data conversion treatment, described data conversion treatment is separated into luminance component view data and color component view data with input image data, and described image processing equipment is carried out the pixel value treatment for correcting according to the translation data that is produced by described color conversion cell.
In addition, an embodiment according to image processing equipment of the present invention, false colour element surveyed area is provided with the unit and has the benchmark image analytic unit, be used to receive benchmark image, and obtain brightness value with the benchmark image pixel that the white saturated pixel in the processed image is corresponding, in described benchmark image, do not have saturated brightness value with the pixel of position that the white saturated pixel in the processed image is corresponding.Described benchmark image is and the photographic images that processed image object is identical.False colour element surveyed area is provided with the unit and also has false colour element surveyed area determining unit, is used for according to the brightness value of the benchmark image pixel of correspondence false colour element surveyed area being set, and described brightness value is obtained by the benchmark image analytic unit.
In addition, an embodiment according to image processing equipment of the present invention, false colour element surveyed area determining unit is provided with false colour element surveyed area by using look-up table, in described look-up table, the brightness value of benchmark image and can produce false colored pixel coverage or the false colour element surveyed area connection that is relative to each other.
In addition, according to an embodiment of image processing equipment of the present invention, false colour element surveyed area determining unit is carried out the processing that wideer false colour element surveyed area is set along with the increase of the brightness value of the respective pixel in benchmark image.
In addition, an embodiment according to image processing equipment of the present invention, be used as benchmark image by using by captured image of the image-capture device with muting sensitivity transducer or the image of taking under the condition of the low exposure that is lower than suitable exposure, false colour element surveyed area determining unit obtains and benchmark image brightness value in the part that the white saturated pixel in the processed image is corresponding.
In addition, a second aspect of the present invention provides a kind of image processing method.Described image processing method comprises:
White saturation detection step detects white saturated pixel from view data;
False colour element surveyed area is provided with step, is arranged on the white saturated pixel false colour element surveyed area on every side that detects in the white saturation detection step;
The false colored step that detects is provided with in the zone that is provided with in the step at false colour element surveyed area, and the pixel that will have the color corresponding with false colour is identified as false colour element; And
The pixel value aligning step is according to the value of surrounding pixel, to carrying out the pixel value treatment for correcting in the false colored false colour element that detects the step detection.
In addition, according to one embodiment of the present of invention, the pixel value aligning step comprises: the compensated interpolation step, and false colour element is carried out compensated interpolation handle according to pixel value on every side; And color Fuzzy Processing step, false colour element is carried out the color Fuzzy Processing.
In addition, an embodiment according to image processing method of the present invention, the pixel value aligning step is such step: select the pixel except false colour element and white saturated pixel around false colour element, and carry out compensated interpolation according to selected pixel and handle from the pixel that exists.
In addition, an embodiment according to image processing method of the present invention, false colored detection step is such step: carry out being provided with in the zone that is provided with in the step at false colour element surveyed area, the pixel that will have default specific color is identified as the processing of false colour element.
In addition, an embodiment according to image processing method of the present invention, described false colour is a purple fringe, and false colored to detect step be such step: carry out the processing that purple pixel in the zone that is provided with in the step is identified as false colour element is set at false colour element surveyed area.
In addition, an embodiment according to image processing method of the present invention, it is such step that described false colour element surveyed area is provided with step: according to take with the aperture during the processed view data, focus information and from light center to white saturated pixel (x, among the data of distance y) at least one carried out the processing of the false colour element surveyed area that is provided with around the white saturated pixel of determining to detect in white saturation detection step.
In addition, according to an embodiment of image processing method of the present invention, described white saturation detection step is such step: carry out the processing that selection has the white saturated pixel of the brightness that is greater than or equal to predetermined threshold.
In addition, an embodiment according to image processing method of the present invention, image processing method also comprises the color conversion step of carrying out data conversion treatment, described data conversion treatment is separated into luminance component view data and color component view data with input image data, and described image processing method also comprises according to carrying out the pixel value treatment for correcting by the translation data of color conversion cell generation.
In addition, an embodiment according to image processing method of the present invention, false colour element surveyed area is provided with step to have: the benchmark image analytical procedure, the input reference image, and obtain brightness value with the benchmark image pixel that the white saturated pixel in the processed image is corresponding, in described benchmark image, do not have saturated brightness value with the pixel of position that the white saturated pixel in the processed image is corresponding, described benchmark image is and the photographic images that processed image object is identical; And false colour element surveyed area determining step, according to the brightness value of the benchmark image pixel of correspondence false colour element surveyed area being set, described brightness value obtains in the benchmark image analytical procedure.
In addition, an embodiment according to image processing method of the present invention, false colour element surveyed area determining step is such step: by using look-up table false colour element surveyed area is set, in described look-up table, the brightness value of benchmark image and can produce false colored pixel coverage or the false colour element surveyed area connection that is relative to each other.
In addition, according to an embodiment of image processing method of the present invention, false colour element surveyed area determining step is such step: execution is provided with the processing of wideer false colour element surveyed area along with the increase of the brightness value of the respective pixel in benchmark image.
In addition, an embodiment according to image processing method of the present invention, in false colour element surveyed area determining step, use is used as benchmark image by captured image of the image-capture device with muting sensitivity transducer or the image of taking under the condition of the low exposure that is lower than suitable exposure, obtain and benchmark image brightness value in the part that the white saturated pixel in the processed image is corresponding.
In addition, a third aspect of the present invention provides a kind of computer program, is used to make the computer carries out image processing.Described program comprises:
White saturation detection step detects white saturated pixel from view data;
False colour element surveyed area is provided with step, is arranged on the white saturated pixel false colour element surveyed area on every side that detects in the white saturation detection step;
The false colored step that detects is provided with in the zone that is provided with in the step at false colour element surveyed area, and the pixel that will have the color corresponding with false colour is identified as false colour element; And
The pixel value aligning step is according to the value of surrounding pixel, to carrying out the pixel value treatment for correcting in the false colored false colour element that detects the step detection.
In addition, according to one embodiment of the present of invention, false colour element surveyed area is provided with step and comprises: the benchmark image analytical procedure, the input reference image, and obtain brightness value with the benchmark image pixel that the white saturated pixel in the processed image is corresponding, in described benchmark image, do not have saturated brightness value with the pixel of position that the white saturated pixel in the processed image is corresponding, described benchmark image is and the photographic images that processed image object is identical; And false colour element surveyed area determining step, according to the brightness value of the benchmark image pixel of correspondence false colour element surveyed area being set, described brightness value obtains in the benchmark image analytical procedure.
Computer program of the present invention can be provided to for example general-purpose computing system, and described general-purpose computing system can be via the storage medium such as CD, FD or MO of computer-readable format or by carrying out various program codes such as the communication media of network.With computer-readable format provide such program can so that computer realize according to described program handling.
By based on the more detailed description according to following embodiment of the present invention and accompanying drawing, other purpose of the present invention, feature and advantage will become clear.Term herein " system " refers to the logical combination of a plurality of equipment, is included in system in the same shell but be not limited to each equipment.
According to configuration of the present invention, detect white saturated pixel from view data, false colour element surveyed area is set around the white saturated pixel that is detected, detect pixel with color corresponding with false colour such as purple fringe from set zone, the pixel that is detected is confirmed as false colour element, and the false colour element of being discerned is carried out treatment for correcting based on the value of surrounding pixel.Therefore, can detect the contiguous false colored zone that produces effectively such as purple fringe in white saturated pixel, and the value of correction pixels partly.Also can produce with the view data of outputting high quality and do not influence entire image.
According to configuration of the present invention, can carry out suitable correction to the false colour such as purple fringe that chromatic aberation causes, described chromatic aberation takes place for the image of being taken by video camera, and can realize producing and the outputting high quality view data.Use traditional video camera, might shoot factitious image, unless under the situation that produces purple fringe, adjust the setting such as lens stop.But, when application is of the present invention, can extracts and proofread and correct the false colour such as purple fringe that in captured image, comprises effectively, and can produce and the outputting high quality image.Needn't pay close attention to lens stop and focal length, purple fringe during taking, not occur, therefore make and to carry out shooting with the higher degree of freedom.
In addition, according to configuration of the present invention, estimate and the white corresponding object intrinsic brilliance of saturated pixel according to benchmark image, false colored (purple fringe) surveyed area corresponding with the object intrinsic brilliance is set, and detect and proofread and correct false colored (purple fringe) pixel in set zone.Therefore, can be provided with and regional corresponding false colored (purple fringe) surveyed area that wherein might produce false colored (purple fringe), and not make set false colored (purple fringe) surveyed area become too big or too little.Can be reliably and detect false colored (purple fringe) pixel effectively, improve the accuracy of image rectification, and raise the efficiency.
Description of drawings
Fig. 1 is the figure that the configuration of image processing equipment of the present invention is shown.
Fig. 2 is the flow chart of the entire process sequence of diagram image processing method of the present invention.
Fig. 3 is the figure of the configuration of the view data that will be proofreaied and correct by image processing equipment of the present invention of diagram.
Fig. 4 is that the detailed sequence of handling is detected in the white zone of saturation of carrying out in image processing of the present invention.
Fig. 5 is the flow chart that is illustrated in the detailed sequence of the false colour element surveyed area set handling of carrying out in the image processing of the present invention.
Fig. 6 is the figure that is illustrated in the processing example of the false colour element surveyed area set handling of carrying out in the image processing of the present invention.
Fig. 7 is illustrated in the flow chart that the false colour element of carrying out in the image processing of the present invention detects the detailed sequence in handling.
Fig. 8 is illustrated in the false colour element of carrying out in the image processing of the present invention to detect processing example (the use CIE L that handles
*a
*b
*Situation) figure.
Fig. 9 is illustrated in the figure that the false colour element of carrying out in the image processing of the present invention detects the processing example of handling (using the situation of YCbCr).
Figure 10 (A) and 10 (B) comprise the figure of the processing example that is illustrated in the false colour element treatment for correcting of carrying out in the image processing of the present invention.
Figure 11 is the flow chart that is illustrated in the detailed sequence of the false colour element treatment for correcting of carrying out in the image processing of the present invention.
Figure 12 is the flow chart that is illustrated in the detailed sequence of the false colour element treatment for correcting of carrying out in the image processing of the present invention.
Figure 13 is the flow chart that is illustrated in the detailed sequence of the color Fuzzy Processing of carrying out in the image processing of the present invention.
Figure 14 be illustrated in the image processing equipment of the present invention, according to the block diagram of functional configuration of Digital Signal Processing that is used to carry out false colour correction of first embodiment.
Figure 15 (a) and 15 (b) comprise the figure of the relation between the generation of configuration of diagram object intrinsic brilliance/edge and false colour (purple fringe).
Figure 16 (a)～(c) comprises the figure of the generation of diagram object intrinsic brilliance, edge configuration, captured image brightness and false colour (purple fringe).
Figure 17 (a) and (b), (p), (r) and (q) comprise that diagram uses the figure of the monochrome information of the benchmark image that muting sensitivity transducer for example obtains to estimate the object intrinsic brilliance.
Figure 18 comprises the figure of the relation between the generation scope that is illustrated in object intrinsic brilliance and false colored (purple fringe).
Figure 19 is the flow chart that the processing procedure of false colored surveyed area is set diagram being used in a second embodiment.
Figure 20 is the form of example of structure that is used for being provided with the look-up table (LUT) that the processing of false colored surveyed area uses that illustrates in a second embodiment.
Figure 21 be illustrated in the image processing equipment of the present invention, according to the block diagram of functional configuration of Digital Signal Processing that is used to carry out false colour correction of second embodiment.
Figure 22 is illustrated in block diagram in the image processing equipment of the present invention, that the configuration of unit is set according to the false colour element surveyed area of Digital Signal Processing that is used for carrying out false colour correction of second embodiment.
Embodiment
Below, with reference to accompanying drawing details according to image processing equipment of the present invention, image processing method and computer program is described.
[first embodiment]
The example of the configuration of image processing equipment at first, is described with reference to Fig. 1.Image processing equipment shown in Figure 1 has image capturing unit, and is illustrated as the example that is used for the data of the image of being caught by image capturing unit are carried out the equipment of treatment for correcting.But for example, image processing equipment of the present invention may be received in the view data of storing in the memory cell such as hard disk, and proofreaies and correct the image of being imported.Described image processing equipment can be not only to carrying out treatment for correcting via the view data of image capturing unit input, and can carry out treatment for correcting to the various types of input image datas such as the view data of importing via storage device or network.Fig. 1 shows an example of the configuration of image processing equipment of the present invention.
The detailed configuration of the image processing equipment that explanation is shown in Figure 1.As shown in Figure 1, described image processing equipment comprises lens 101, aperture 102, solid-state image capture device 103, correlated double sampling circuit 104, modulus (A/D) transducer 105, DSP piece 106, timing generator 107, digital-to-analogue (D/A) transducer 108, video encoder 109, video-frequency monitor 110, coding decoder (CODEC) 111, memory 112, CPU 113, entering apparatus 114, quickflashing control device 115 and quickflashing luminescent device 116.
At this, the action button such as record button that entering apparatus 114 fingers provide on the video camera master unit etc.DSP piece 106 is the pieces with signal processor and image RAM.Described signal processor can be carried out the image processing of programming in advance to the view data of storing in image RAM.Below, the DSP piece will be called " DSP " for short.
The following describes the integrated operation of present embodiment.
The incident light that has arrived solid-state image capture device 103 by optical system at first arrives the light receiving element on the image capture surface, and is the signal of telecommunication by the light receiving element opto-electronic conversion.The noise of the described signal of telecommunication is eliminated by correlated double sampling circuit 104, and the signal that as a result of produces is converted to digital signal by analog to digital converter 105.Thereafter, digital signal is temporarily stored in the video memory in the digital signal processor (DSP) 106.During taking, can be so that quickflashing luminescent device 116 be luminous via quickflashing control device 115 where necessary.
When catching image, timing generator 107 control signal treatment systems remain on the constant frame rate so that image caught.Also send pixel stream, in digital signal processor (DSP) 106, carry out appropriate image processing to digital signal processor (DSP) 106.Thereafter, the view data that as a result of produces of any one among digital to analog converter 108 and coding decoder (CODEC) 111 or both transmissions.Digital to analog converter 108 is converted to analog signal to the view data that sends from digital signal processor (DSP) 106, and video encoder 109 is a vision signal with described analog signal conversion, and described vision signal can be monitored on video-frequency monitor 110 then.This video-frequency monitor 110 is in the present embodiment as video camera view finder (finder).Coding decoder (CODEC) 111 is the coded image datas that send from digital signal processor (DSP) 106, and with coded image data storage in memory 112.In this case, described memory 112 can be the memory device that for example uses semiconductor, magnetic storage medium, magnetic-optical storage medium or optical storage medium.
The whole system of digital camera system is described as an example of the image processing equipment of present embodiment above.Carry out according to image processing of the present invention by digital signal processor (DSP) 106, promptly be used for the processing of image correcting data.Describe described image processing below in detail.
The feature that how to produce false colored (purple fringe) will at first be described.False colored (purple fringe) refers to that the gamut that wherein causes owing to chromatic aberation significantly appears at white saturated part phenomenon on every side, and has the feature that for example produces the false colour of purple.The size that produces the zone of false colour around white saturated part is associated with the type of optical system, aperture and the focal length and the luminance difference between hi-lite and peripheral part thereof of optical system, and also according to changing with the distance of optical centre.And, on the direction of the outside from the optical centre to the high luminance pixel, first-class to the direction of optical centre from high luminance pixel, a large amount of false colours appears.The optical system that the trend that occurs is caught according to carries out image and difference.According to aforesaid feature, the invention provides a kind of image processing method that is used to reduce purple fringe.Refer to saturated pixel on image at this high luminance pixel, and followingly will be called as white saturated pixel.
As mentioned above, the false colored typical phenomenon that is caused by lens aberration that partly produces in the white saturated high-contrast edges that takes place as the saturated state of intensity level is called as " purple fringe ".Described false colour is not limited to purple, and can be to have green color.In the following embodiments, explanation is produced the false colored example of purple, but the present invention is not restricted to purple with false colour, and is applicable to that correction has the false colored processing at the contiguous any color that produces of white saturated part.
Fig. 2 is the flow chart that present embodiment is shown.To the overview of entire process be described at first, the details of each processing is described then.Step S101 is the color space conversion process, and wherein, input picture is changed through color space, and is separated into luminance component image and color component image.Then, step S102 is that processing is detected in white zone of saturation, wherein, detects white saturated part from the input image data that will proofread and correct.Step S103 is false colour element surveyed area set handling, wherein, the region of search of purple pixel is set, and described purple pixel is the false colour element that produces at white saturated part periphery.And step S104 is that false colour element detects processing, wherein detects the purple pixel in the region of search.Step S105 is false colour element treatment for correcting, wherein proofreaies and correct the color component of the purple pixel that is detected.And step S106 is color Fuzzy Processing (filtration treatment), wherein, carries out filtration treatment so that the result who seems nature to be provided.At last, step S107 is that the color space inverse conversion is handled, wherein, the result's that conversion is proofreaied and correct luminance component image and color component image, and the image that will as a result of produce is output as the RGB coloured image.
Input picture is the data of the image of being caught by for example image capturing unit in the system with image capturing unit shown in Figure 1, and is to be imported into the data of digital signal processor (DSP) 106 as numerical data.As mentioned above, input image data can be view data of storing in storage medium or the data of importing from another equipment by network.
In the present embodiment, processed input picture to be called as RGB
InRGB
InBe image, and be to go mosaic (demosaic) processing and white balance processing to obtain by the data of exporting from the solid-state image capture device of video camera are carried out with the color information in each pixel.With processed input picture RGB
In, as shown in Figure 3, at input picture RGB
InThe x direction on the quantity (width) of pixel be indicated as w, and the quantity of the pixel on the y direction (highly) is indicated as h.Describe each step below in detail.
At first, in the color space conversion process in step S101, input picture RGB
InBe converted, and be separated into luminance component image L
InWith the color component image C
InThat is, the RGB coloured image of input is converted into color space data with luminance component and the color space data with color component.For example, YCbCr or CIEL
*a
*b
*The color space that can be used for institute's translation data.
Then, during the white zone of saturation in step S102 is detected and handled, carry out based on the white zone of saturation of the view data of being changed and detect processing (white saturated shielding is determined).That is, detect white saturated pixel, and produce white saturated shielding S as binary picture from view data.Illustrate that with reference to Fig. 4 the white saturated pixel of carrying out detects the details of handling in step S102.
In step S201 and S202, in the view data that will proofread and correct, the locations of pixels that initialization will be checked.In described initialization, the locations of pixels that check is set to x=0 and y=0.In step S203, determine to have luminance component image L
InPixel (x, whether brightness y) is greater than or equal to predetermined threshold.When described brightness is greater than or equal to described threshold value, handle and proceed to step S204, wherein, (x y) is set to very with white saturated shielding S.When threshold value is less than or equal in brightness, handle and proceed to step S205, wherein, (x y) is set to vacation with white saturated shielding S.
At step S206, upgrade location of pixels x.At step S207, determine whether that x reaches maximum (x=w).When x does not reach maximum, repeated execution of steps S203 and step subsequently.When x in step S207 reaches maximum (x=w), handle proceeding to step S208, upgrade location of pixels y.At step S209, determine whether that y reaches maximum (y=h).When y does not reach maximum, repeated execution of steps S202 and step subsequently.When the processing that is through with for all pixels (x=0 is to w, and y=0 is to h), finish to produce white saturated shielding.
As a result, each location of pixels is created in the pixel region with the intensity level that is higher than threshold value and has the white saturated shielding of distinguishing between the pixel region of luminance area of the threshold value of being equal to or less than.When being provided with the intensity level of 0 (secretly)-255 (bright), described threshold value is set to for example about 250, and will have the pixel extraction that is greater than or equal to 250 intensity level is white saturated generation pixel, and produces the white saturated shielding that only can distinguish those pixels.
By aforesaid processing, step S102 shown in Figure 2 is through with.Then, the false colour element surveyed area set handling of execution in step S103.In this is handled, carry out for (expansion (dilate)) being set and handle in the colored region of search that produces part of the vacation of white saturated part periphery.In this case, produce and to be used to the false colored shielding P that produces the region of search of part that distinguishes at white saturated part periphery, promptly represent shielding P in the region of search of the purple pixel of white saturated part periphery.P is the binary masks image.
The details of the false colored generation part region of search set handling in step S103 is described with reference to Fig. 5.At first, in step S301, all pixels that false colored region shielded P are initialized as vacation.Then, in step S302 and S303, initialized pixel location.In described initialization, location of pixels is set to x=0 and y=0.
After being provided with location of pixels,, determine the pixel value of the white saturated shielding S of set location of pixels at step S304.(x, y) (x y) is fictitious time, handles and proceeds to step S314 to carry out the processing to next pixel for set location of pixels as white saturated shielding S.If (x is true y) to S, then handles proceeding to step S305.(x, y) (x is a fictitious time y), this means location of pixels (x, the non-white saturated pixel of pixel value indication y) for location of pixels as white saturated shielding S.(x, y) (x is a true time y), this means location of pixels (x, the white saturated pixel of pixel value indication y) for location of pixels as white saturated shielding S.
(x, y) (x y) is true time, and at step S305, determining wherein can be in white saturated pixel (x, y) the false colored zone of periphery generation for location of pixels as white saturated shielding S.As mentioned above, purple fringe occurs, and the zone of purple fringe wherein can occur and change according to the type of optical system, the setting and the luminance difference between high luminance pixel and surrounding pixel thereof of aperture/focal length at white saturated part periphery.The zone that purple fringe wherein can take place also changes according to the distance of the optical centre on the distance image.Can be during taking the type of storing optical system and the information of setting, to be used for image processing.Can easily determine the position relation between white saturated part and optical centre part.
For example, according to the feature of the optical system of image capture system, use default false colour element surveyed area to determine information in the present embodiment, such as the look-up table (LUT) that comprises corresponding to the false colour element surveyed area of image-region.For example, described look-up table (LUT) has the structure that can export as the false colored surveyed area information of input parameter, the colored surveyed area information of described vacation is such as (x, distance y), the aperture during the shooting of view data and focus information from the optical centre on the image to white saturated pixel.
Aperture and the focus information of attribute information acquisition during the shooting of view data from view data, and the described location of pixels information that is set to the information that will check in this flow process is used for (x, distance y) from the optical centre on the image to white saturated pixel.
In the time can not obtaining aperture during shooting and the attribute information the focus information in view data, can be only according to (x, distance y) is determined false colour element surveyed area from the optical centre on the image to white saturated pixel.And, can (x, y) Zhou Bian certain zone be set to false colour element surveyed area automatically, and does not use such as (x, the information of distance y) and so on from optical centre to white saturated pixel white saturated pixel.
In step S305, (x, y) scope of Zhou Bian false colour element surveyed area is defined as four scalar value (x with white saturated pixel
0, x
1, y
0, y
1).In this is determined, as shown in Figure 6, will be by saturated white pixel (x, y) x of 201 peripheries
0To x
1And y
0To y
1The rectangular area that limits is defined as false colour element surveyed area 202.By the pixel 203 of the dotted line among Fig. 6 indication is except the white saturated pixel paid close attention to (x, y) the white saturated pixel 203 outside 201.
Then, about determined false colour element surveyed area (x
0, x
1, y
0, y
1), will be by i=x
0And j=y
0The point of expressing is set to initial point (at step S306 and S307).In step S308 to S312, about at false colour element surveyed area (x
0, x
1, y
0, y
1) in each pixel and determine whether that (i, value j) is a true or false to white saturated shielding S.About the pixel that comprises in the false colour element surveyed area 202 shown in Fig. 6, the value that determines whether white saturated shielding in regular turn is a true or false.In Fig. 6, about the white saturated pixel paid close attention to (x, y) 201 and white saturated pixel 203, determine that (i is true j) to white saturated shielding S.
When determining that in step S308 (i j) is true time to white saturated shielding S, handles and proceeds to step S310, wherein carries out the processing for next pixel.(i j) is fictitious time, handles and proceeds to step S309, and wherein, (i, value j) is set to very with false colored region shielding P as white saturated shielding S., handle proceed to step S310, wherein, carry out processing next pixel thereafter.Step S310-S313 comprises the processing of the value that is used to upgrade i and j and is used to determine whether to satisfy maximum i=x
1And j=y
1Processing.When upgrading the value of i and j in regular turn, determine in specific pixel (x, y) false colored region shielding P (i, value j) of periphery setting.
False colored shielding P (i, j)=the described pixel of true respresentation belongs to false colour element surveyed area, and is not white saturated pixel.False colored region shielding P (i, j)=the described pixel of false expression does not belong to false colour element surveyed area, or white saturated pixel.
When in step S313, determining to finish at false colour element surveyed area (x
o, x
1, y
0, y
1) in all pixels (i, during j) processing, described processing proceeds to step S314.Step S314-S317 comprises the processing of the value of upgrading x and y and determines whether to satisfy the processing of maximum x=w and y=j.When upgrading the value of x and y in regular turn, about all pixels in the processed image (x=0 to w and y=0 to h) are determined false colored region shielding P (i, value j), and produce false colored region shielding P (i, j).When determining that (x in the time of y), finishes the false colour element surveyed area set handling in the step S103 shown in Fig. 2 to treated all pixels in described image.
Next processing is that the false colour element in the step S104 shown in Fig. 2 detects processing.In step S103, determined the false colored of white saturated part periphery though use false colored region shielding P, promptly wherein can in image, produce the zone of purple fringe, but in next step S104, detect the processing that in fact is assumed to be pixel with purple fringe.That is, from the pixel during satisfying the genuine zone of false colored region shielding P=, detect the pixel that in fact supposition has purple fringe.
In step S104, the false colored region shielding P by overwrite calculates in step S103 upgrades false colored region shielding P, produces the false colored region shielding P that makes it possible to only discern the pixel that finally is defined as having false colour thus.Illustrate that with reference to Fig. 7 the false colour element in step S104 detects the details of handling.
In step S401 and S402, initialisation image data pixels position.In described initialization, location of pixels is set to x=0 and y=0.In step S403, determine location of pixels (x, the value of false colored region shielding P y).(x y)=vacation, handles proceeding to step S406 for P.(x y)=very, handles proceeding to step S404 for P.
As mentioned above, and false colored region shielding P (i, j)=the described pixel of true respresentation belongs to false colour element surveyed area, and is not white saturated pixel.False colored region shielding P (i, j)=the described pixel of false expression does not belong to false colour element surveyed area, or white saturated pixel.
(x, (i, j)=true time, promptly described pixel belongs to false colour element surveyed area and when not being white saturated pixel, determines whether (x, color component image C y) in the position at step S404 y) to satisfy false colored region shielding P when location of pixels
InColor be corresponding to false colored color.For example, determine whether that described color is the purple of purple fringe.
Determine whether the color component image C
In(x is to change according to the color space of being changed with the method for the false colored corresponding color of for example purple y).With reference to Fig. 8 for example CIE L is described
*a
*b
*An example of the definite method when being used for color space.Use component a
*With component b
*Determine color component.CIE L
*a
*b
*Has following feature.As component a
*Value when increasing, color is near red (Red), as component a
*Value when reducing, color is near green (Green).As component b
*Value when increasing, color is near yellow (Yellow), as component b
*Value when reducing, color is near blue (Blue).Therefore, when in four-quadrant, having (a
*, b
*) time, it has the color near purple (Purple).Use is by vector (a
*, b
*) and axle a
*The angle [alpha] that forms determines whether that color is a purple.
The angular range that is confirmed as the purple of false colour suitably is set to parameter.Perhaps, predetermined value is used for described angular range.The saturated trend of the color of false colored (purple fringe) partial pixel is high.Therefore, only working as above-mentioned condition is satisfied and color component C
InColor saturated when being greater than or equal to predetermined threshold value, can be defined as false colored (purple fringe).
Determine whether the color component image C
In(x is to change according to the color space of being changed with the method for the false colored corresponding color of for example purple y).The following describes the processing that is used for the situation of color space for YCbCr.In this case, use component Cb and component Cr to determine color component.Because Cb=128 and Cr=128 are illustrated in the achromaticity color in the YCbCr space, the color of therefore representing by Cb and Cr be expressed as (Cb-128, Cr-28).For example, by (Cb-128, the angle that θ Cr-128)=(a * cos θ, a * sin θ) expresses is used as tone.By for example determining whether that angle θ determines whether the color component image C in preassigned scope
In(x is corresponding to false colored color y).
Be described as follows processing with reference to Fig. 9: whether determine in the configuration of using YCbCr C
In(x y) has tone in preassigned tone range, promptly has corresponding to the tone in the false colored tone range such as purple fringe (PF).Fig. 9 is the figure that is illustrated in purple fringe (PF) tone range in the color space.In Fig. 9, drawn when the YCbCr space is used as example of color space CbCr two-dimensionally corresponding to color information.The value of trunnion axis indication Cb, the value of vertical axis indication Cr.Initial point 250 has the color corresponding to Cb=128 and Cr=128.
Method as specifying corresponding to the tone range of purple fringe (PF) is used for determining whether the method for described tone in the scope of two color tones.Described purple fringe has purple, and in two-dimentional CbCr coordinate space shown in Figure 9 this specific tone range is set.Zone between tone line shown in the described accompanying drawing 251 and tone line 252 is set to the zone of indication corresponding to the tone range 253 of purple fringe (PF).
Can be according in tone range shown in Figure 9 253, whether comprising C
In(x y) determines whether C
In(x is corresponding to false colored color y).
The angular range that is confirmed as the purple of false colour suitably is set to parameter.Perhaps, predetermined value is used for described angular range.The saturated trend of the color of false colored (purple fringe) partial pixel is high.Therefore, only ought satisfy above-mentioned condition and color component C
InColor saturated when being greater than or equal to predetermined threshold value, it can be defined as false colored (purple fringe).By this way, as definite C
In(x when y) being vacation colour (purple fringe) pixel, handles proceeding to step S406.As definite C
In(x when y) not being false colored (purple fringe), handles proceeding to step S405.
In step S405, (x y) is set to vacation to false colored region shielding P, and processing proceeds to step S406.That is, be confirmed as to have false colour in the above-mentioned processing in step S103 (Fig. 2), promptly false colored region shielding P (x, y)=very, pixel finally is defined as not having false colour, according to this final determining, change (renewals) be false colored region shield P (x, y)=vacation.
Promptly, in processing shown in Figure 7, only determined finally to be defined as having false colored pixel by described color, classified from being defined as by the above-mentioned processing the step 103 (Fig. 2) may having in the false colored pixel, and upgrade false colored region shielding P (x, y).
Step S406 to S409 comprises the processing of the value of upgrading x and y and determines whether to satisfy the processing of maximum x=w and y=h.When upgrading the value of x and y in regular turn, about all pixels in the processed image (x=0-w and y=0-h) are determined false colored region shielding P (i, value j), and produce false colored region shielding P (i, j).(x in the time of y), finishes false colour element in the step S104 shown in Fig. 2 and detects and handle when determining all pixels in image treated.
The following describes the false colour element treatment for correcting among the step S105 shown in Figure 2.In step S104, as mentioned above, produced and be used for the false colored region shielding P that identification is confirmed as having the pixel of false colored (purple fringe).In next step S105, compensate reprocessing by pixel value, and carry out the processing of color that interpolation is confirmed as having the pixel of false colored (purple fringe) based on the pixel around the pixel.Handle in this " compensation " and to refer to repeat false colour element interpolation in regular turn and handle according to the pixel value of surrounding pixel.
Figure 10 (A) and 10 (B) are the schematic diagrames that is illustrated in the false colour element treatment for correcting among the step S105 shown in Fig. 2.Shown in Figure 10 (A), be classified into any one in " white saturated pixel 301 ", " false colored (purple fringe) pixel 302 " and " neither pixel 303 of white saturated non-again purple fringe pixel " in all pixels in the processing of step S104.
In the first step of compensation reprocessing, scanning each pixel in image.If false colored (purple fringe) pixel 302 (position (x, y)) neither the vicinity of the pixel 303 of white saturated non-again purple fringe pixel, for example, in 8 neighborhood pixels around if pixel 303 is present in, then the mean value of the color components in pixels C of neighborhood pixels 303 be set to new color components in pixels C (x, y).
By this way, all pixels are carried out the processing that the pixel value of false colored (purple fringe) pixel 302 once is set according to other pixel 303.As a result, shown in Figure 10 (B), the color component of false colored (purple fringe) pixel 302 of other pixel 303 vicinities is interpolated, and colored (purple fringe) pixel 302 of vacation is set to colored (purple fringe) pixel 304 of vacation of interpolation.
In addition, in the compensation reprocessing below, come the color component of false colored (purple fringe) pixel 302 of vicinity of false colored (purple fringe) pixel 304 of other pixel 303 of interpolation similarly or interpolation according to false colored (purple fringe) pixel 304 of other pixel 303 and interpolation.Such reprocessing is repeated fixed number of times.In this was handled, false colored (purple fringe) pixel 302 was set to false colored (purple fringe) pixel 304 of interpolation on by arrow 320 indicated directions shown in Figure 10 (B) in regular turn.After the such reprocessing that has repeated fixed number of times, even false colored (purple fringe) pixel 302 that exists its color component also not to be interpolated, determine that also it is impossible using the color interpolation of their pixels on every side, and carry out the saturated processing of color of the color that reduces those pixels equably.
By this way, false colour element treatment for correcting among the step S105 shown in Figure 2 is the compensation reprocessing, promptly is arranged on the processing of the pixel value of the pixel that is confirmed as having false colored (purple fringe) among the step S104 according to the pixel except around the white saturated pixel.Remaining false colored (purple fringe) pixel 302 that is not interpolated by above-mentioned processing is carried out the saturated reduction of color to be handled.
The details of the false colour element treatment for correcting in step S105 is described with reference to the handling process shown in Figure 11 and 12.In the handling process in Figure 11 and 12, processing in the step S501-S515 shown in Figure 11 is the processing that is used for the color component of compensatory time-off colour (purple fringe) pixel, and the processing in the step S516-S523 shown in Figure 12 is to be used to reduce the saturated processing of the color of the pixel that is not also compensated.
Describe in regular turn from Figure 11.At first, in step S501, the content of the false colored region shielding P that the false colour element in the step S104 shown in Fig. 2 produces in detecting and handling is copied to the binary masks P with roughly the same size " on.Example comprises:
False colored region shielding P=is true → binary masks P "=true (1)
False colored region shielding P=vacation → binary masks P "=false (0)
At step S502, the value of the variable t of the number of repetition of indication compensation deals is set to 0.In step S503, binary masks P " value be copied on another binary masks P ' with same size.These shield P " and P ' be the make-up shielding image that is used to store the position of false colored (purple fringe) pixel 302 that reduces by reprocessing.
Then, in step S504 and S505, indication is initialised the variable x and the y of processed locations of pixels coordinate to satisfy x=0 and y=0.Then, in step S506, determine binary masks P ' (x, value y).(x y)=vacation, that is, when processed pixel is not false colour element, handles proceeding to step S510 for binary masks P '.(x y)=very, promptly when processed pixel is false colour element, handles proceeding to step S507 for binary masks P '.
In step S507, determine whether that (there is the pixel that satisfies binary masks P '=vacation and white saturated shielding S=vacation in x in the pixel of vicinity y) in pixel.That is, determine whether to exist neither the false colour element pixel of non-white saturated pixel again.When determining that (x when not having the pixel that satisfies such condition in the pixel of vicinity y), handles proceeding to step S510 in pixel.
When determining that (x exists in the pixel of vicinity y) neither false colour element again during the pixel of non-white saturated pixel, handles proceeding to step S508 in pixel.At step S508, determine and for coordinate (x, y) (x y) is provided with and satisfies binary masks P '=vacation and white saturated shielding S=vacation color component C, promptly neither the false colour element mean value of the color component C of one or more pixels of non-white saturated pixel again.That is, determined mean value is set to pixel (x, pixel value y).Handle by this interpolation, be arranged on false colored (purple fringe) pixel 304 of the interpolation shown in Figure 10 (b).
After this is handled, handle proceeding to step S509.At step S509, with binary masks P " (x, value y) is set to vacation.That is, at binary masks P " (x, y) on, the pixel of changing into the false colour element of interpolation from false colour element is set to be identified as the pixel of non-false colour element.
Step S510-S513 comprises processing and the definite peaked processing of renewal with the coordinate of processed pixel.At step S510, x is sat target value increase 1.Then, at step S511, determine whether to satisfy x＞w-1 (w is the width of input picture).When determining to satisfy x＞w-1, handle proceeding to step S512.When determining not satisfy x＞w-1, handle proceeding to step S506.
At step S512, the y coordinate increases 1.Then, at step S513, the y coordinate is compared with h-1, h-1 is the maximum (h is the height of input picture) of y coordinate.When determining to satisfy y＞h-1, handle proceeding to step S514.When determining not satisfy y＞h-1, handle proceeding to step S505.
At step S514, the t of indication number of repetition increases 1.At step S515, determine whether that number of repetition t is predetermined value tmax or bigger.Simultaneously, for masked images P ", the quantity that determines whether to have the pixel of true value is 0.That is, determining whether to handle by compensated interpolation is not 0 by the quantity of the false colour element of complete interpolation also.When determine above-mentioned two conditions any one be true time, processing proceeding to step S516.When definite two conditions are fictitious time, handle and return step S503, and carry out compensation deals once more.
At step S515, when determining whether that number of repetition t reaches in predetermined value tmax or the bigger processing, not not in 0 the processing perhaps, be defined as true time, handle proceeding to step S516 shown in Figure 12 determining whether to handle also by the quantity of the false colour element of interpolation fully by compensated interpolation.
In step S516 and S517, indication is initialised the variable x and the y of the coordinate of processed pixel to satisfy x=0 and y=0.Then, in step S518, determine whether (x, binary masks P y) " value be true.That is, determine whether that with processed pixel be false colour element.For binary masks P " (x, y)=true, promptly for false colour element, handle proceeding to step S519.For binary masks P " (x, y)=vacation, promptly for non-false colour element, handle proceeding to step S520.
When processed pixel is false colour element, in step S519, carry out the saturated reduction of color and handle, (x, (x, color y) is saturated for color component C y) to be used to reduce pixel.For example, when using L
*a
*b
*During color system, can pass through a
*And b
*Each component on duty to reduce color equably with the constant in scope 1.0 to 0.0 saturated.When the saturated reduction processing of the color in step S519 finishes, handle proceeding to step S520.At step S520, x is sat target value increase 1.Then, at step S521, determine whether to satisfy x＞w-1.When determining to satisfy x＞w-1, handle proceeding to step S522.When determining not satisfy x＞w-1, handle proceeding to step S518, and the neighborhood pixels that has upgraded the x coordinate is repeated similarly to handle.
At step S521, when determining to satisfy x＞w-1, reach the maximum of x coordinate, therefore handle and proceed to step S522, wherein the y coordinate is increased 1.At next step S523, the y coordinate is compared with h-1, h-1 is the maximum of y coordinate.When not satisfying y＞h-1, handle proceeding to step S517, and the pixel of y coordinate with renewal is repeated similarly to handle.When determining to satisfy y＞h-1 at step S523, finish false colour element treatment for correcting.By aforesaid processing, the false colour element treatment for correcting among the step S105 shown in Figure 2 finishes.
When the false colour element treatment for correcting that finishes in step S105, then carry out the color Fuzzy Processing (filtration) in step S106.At step S105, the color of false colored (purple fringe) pixel of interpolation by repeating compensation deals, and false colored (purple fringe) pixel that does not have to compensate is through reducing the saturated processing of color.But its colored part that changes quickly relatively may reside in false colored (purple fringe) part of the processing of process in step S105.In the color Fuzzy Processing (filtration) in step S106, carry out the processing of using blur filter and change to reduce quick color.
The details of the color Fuzzy Processing (filtration) in step S106 is described with reference to handling process shown in Figure 13.At first, at step S601, indicate the value of variable t of the number of repetition of color Fuzzy Processing (filtration) to be set to 0.Then, in step S602, the color component image C of being upgraded by the false colour element treatment for correcting among the aforesaid step S105 is copied to the color component image C with roughly the same size ' on.
Then, in step S603 and S604, indication is initialised the variable x and the y of the coordinate of processed pixel to satisfy x=0 and y=0.Then, at step S605, determine false colored region shielding P (x, value y).(x y)=vacation, promptly for non-false colour element, handles proceeding to step S607 for false colored region shielding P.(x y)=very, promptly for false colour element, handles proceeding to step S606 for false colored region shielding P.
False colored region shielding P (x, y) be to be set to really shield for following pixel, it is colored that described pixel determines to be confirmed as having vacation according to the color (pixel value) in the false colour element detection processing (corresponding to the handling process among Fig. 7) in aforesaid step S104.(x is for the false colour element zone of being proofreaied and correct by the false colour element correction pixels processing among the aforesaid step S105 and the shadow data of retention value " very " y) to false colored region shielding P.Because false colored region shielding P (x y) is used for color fuzzy (filtrations) and handles, and therefore false colored region shielding P (x, duplicating y) shields P " be used for false colour element treatment for correcting at step S105.
Continue the explanation of the color Fuzzy Processing (filtration) among Figure 13.(x y) is true time, promptly when described pixel is false colour element, handles proceeding to step S606 as false colored shielding P in step S605.In step S606, use blur filter to color component.This be surrounding pixel as indicated color component pixel value C and upgrade indication in the color component image C ' in pixel (x, the pixel value C of color component y) (x, the processing of value y).For example, determine to comprise that (x y) and the mean value of the color component C ' of 9 pixels of 8 pixels on every side, and uses the rolling average filter so that described mean value is set to processed pixel (x with processed pixel, the pixel value C of renewal y) (x, y).
After blur filter in end step S606 is used and handled,, upgrade coordinate, and check maximum processed pixel at step S607-S610.At step S607, x sits target value and increases 1.Then, at step S608, determine whether to satisfy x＞w-1.When determining to satisfy x＞w-1, handle proceeding to step S608.When determining not satisfy x＞w-1, handle and return step S604, and the pixel of upgrading the x coordinate carried out similarly handle, and carry out Fuzzy Processing where necessary.
When determining to satisfy x＞w-1 at step S608, handle and proceed to step S609, wherein, the y coordinate is increased 1.In next step S610, the y coordinate is compared with h-1, h-1 is the maximum of y coordinate.When determining not satisfy y＞h-1, handle proceeding to step S603, and the pixel of y coordinate with renewal carried out similarly handle, and carry out Fuzzy Processing where necessary.
When determining to satisfy y＞h-1 at step S610, handle proceeding to step S611.At step S611, carry out the processing of the number of repetition t of color Fuzzy Processing (filtration) increase 1.In step S612, determine whether that number of repetition t is predetermined threshold t ' max or bigger.For t＜t ' max, handle and return step S602, for t 〉=t ' max, finish color Fuzzy Processing (filtration).
By aforesaid processing, the color Fuzzy Processing (filtration) among the step S106 shown in Figure 2 finishes.When the color Fuzzy Processing (filtrations) that finishes in step S106,, change the luminance component image L of original image at next step S107
InWith the color component image C that produces from above-mentioned processing, and the image RGB that will as a result of produce
OutBe output as the RGB coloured image, finish entire process thus.
Though described the example that false colour is defined as purple in step S404 shown in Figure 7 in above-mentioned processing example, false colored color is not limited to purple, but can be for example to have green color.In this case, be set to different colors, and configuration is used to determine to have false colored region shielding P as the pixel of false colored set color, and (x y) allows processing corresponding to arbitrary hue thus at the false colour of step S404.
When proofreading and correct multiple different color, when for example comprising the multicolour of purple and green, can repeatedly carry out processing shown in Figure 2 to different colors, perhaps can use following configuration to carry out processing shown in Figure 2, in described configuration, the pixel value of the multicolour that comprises purple and green is defined as genuine false colored region shielding P, and (x y) is set to shielding.This makes to proofread and correct and comprises the false colored of multicolour.
Figure 14 is the block diagram that is illustrated in the functional configuration of the digital signal processor (DSP) (corresponding to DSP shown in Figure 1 106) in the image processing equipment of the present invention of carrying out above-mentioned processing.With reference to the block diagram shown in Figure 14, compare with flow chart shown in Figure 2 simultaneously, the processing of being carried out by digital signal processor (DSP) 106 is described.
On function, as shown in figure 14, digital signal processor (DSP) 106 has color conversion cell 401, saturation detection unit 402, false colour element surveyed area are provided with unit 403, false colored (purple fringe) detecting unit 404, pixel value correcting unit 410 and color inverse conversion unit 405 in vain.Pixel value correcting unit 410 comprises false colored (purple fringe) compensated interpolation unit 411 and false colored (purple fringe) Fuzzy Processing unit 412.
Digital signal processor (DSP) 106 receives view data (RGB from input picture frame memory 420
In), and color conversion process unit 401 is the color space conversion of the RGB image of input, and be separated into luminance component and color component.The processing of being carried out by color conversion process unit 401 is corresponding to the processing among the step S101 in the flow chart shown in Fig. 2.Input picture RGB
InBe converted, and be separated into luminance component image L
InWith the color component image C
InThat is, the rgb color input picture is converted into color space with luminance component and the color space with color component.YCbCr or CIE L
*a
*b
*Deng being used for described color space.
Then, according to by color conversion process unit 401 data converted, white saturation detection unit 402 detects has white saturated pixel portion.This is handled corresponding to the step S102 in flow chart shown in Figure 2.On concrete, as above described with reference to flow chart shown in Figure 4, detection has the pixel of the brightness that is greater than or equal to the predetermined luminance threshold value, and carries out white saturated shielding S (x, processing y) that generation is used to discern white saturated pixel.
False colour element surveyed area is provided with unit 403 and carries out the processing that is used for being provided with in the part of the white saturated pixel periphery that is detected by white saturation detection unit 402 false colour element surveyed area.This regional set handling be used for determining wherein can satisfy white saturated shielding S (x, y)=(x, y) periphery produces the processing in the zone of false colour to genuine white saturated pixel, and false colour element surveyed area is determined that information is used for it.This is handled corresponding to the step S103 in flow chart shown in Figure 2, and is performed according to handling process shown in Figure 5.
For example, use have with from the optical centre on the image to white saturated pixel (x, the look-up table (LUT) of the false colour element surveyed area information that distance y) is associated and aperture during captured image data and focus information, false colour element surveyed area is provided with unit 403, and (x, y) periphery is provided with false colour element surveyed area in each white saturated pixel.In the time can not obtaining aperture during captured image data and focus information etc., can be only according to (x, distance y) is determined false colour element surveyed area from the optical centre on the image to white saturated pixel.And (x, y) Zhou Bian certain zone can always be set to false colour element surveyed area, and does not use such as (x, the information of distance y) and so on from optical centre to white saturated pixel in white saturated pixel.
As above described with reference to Fig. 6, false colour element surveyed area is provided with that unit 403 will (x, y) scope of the false colour element surveyed area of 201 peripheries be defined as four scalar value (x in white saturated pixel
0, x
1, y
0, y
1), and further produce the false colored region shielding P that obtains by the white saturated pixel of removing in that zone.
False colored (purple fringe) detecting unit 404 uses by false colour element surveyed area the false colored region shielding P that unit 403 is provided with is set, and further carries out and detect by each pixel being carried out the processing that color determines to handle the pixel that is confirmed as having false colored (purple fringe).That is, upgrade false colored region shielding P to produce the false colored region shielding P that can only discern the false colour element that to proofread and correct.This is handled corresponding to the step S104 in flow chart shown in Figure 2, and is performed according to handling process shown in Figure 7.As mentioned above, be defined as false colored color is arbitrarily, therefore, the purple that wherein only has a specific color value be set to false colored configuration or wherein the multicolour such as green and purple to be set to false colored configuration be possible.
False colour (purple fringe) compensated interpolation unit 411 is carried out and is proofreaied and correct the processing that is defined as having the colored pixel of vacation by false colored (purple fringe) detecting unit 404.Described processing is corresponding to the step S105 in flow chart shown in Figure 2, and is performed according to the handling process shown in Figure 11 and 12.This pixel value treatment for correcting comprises that compensated interpolation is handled (seeing Figure 10 and 11) and (seeing Figure 12) handled in the saturated reduction of color, it is that pixel for the value according to the pixel except false colored and white saturated pixel on every side is confirmed as having false colored (purple fringe) is carried out that described compensated interpolation is handled, and the saturated reduction of described color is handled at the compensated interpolation that does not also have scheduled number of repetition and handled false colored (purple fringe) pixel of proofreading and correct.
The 412 pairs of data of being proofreaied and correct by the processing of false colored (purple fringe) compensated interpolation unit 411 in false colored (purple fringe) Fuzzy Processing unit are carried out Fuzzy Processing.This is handled corresponding to the step S106 in flow chart shown in Figure 2, and is performed according to handling process shown in Figure 13.
False colored (purple fringe) Fuzzy Processing unit 412 extracts by false colored (purple fringe) detecting unit 404 and is defined as having false colored pixel, and uses blur filter to color component.For example, pixel (the x that false colored (purple fringe) Fuzzy Processing unit 412 is determined to comprise with processed, and carry out described mean value and be set to y) and the mean value of 9 pixels of 8 pixels on every side, with processed pixel (x, the rolling average filtration treatment of renewal pixel value y).
The luminance component image L of color inverse conversion unit 405 conversion original images
InWith color component image C as above-mentioned result, and the image RGB that will as a result of produce
OutBe output as the rgb color image.
When using when of the present invention, can to cause by chromatic aberation, carry out suitable correction such as purple fringe false colored, wherein said chromatic aberation takes place the image of being taken by video camera, and can realize the generation and the output of high quality graphic data.Use traditional video camera, might take factitious image, unless under the situation that produces purple fringe, adjust setting such as lens stop.But, when using image processing of the present invention, can carry out and proofread and correct so that captured image seems more natural.Therefore, needn't pay close attention to lens stop and focal length, purple fringe does not take place during taking, therefore make and to take with the higher degree of freedom.
[second embodiment]
Then, will the configuration of the second embodiment of the present invention be described.As describing in detail with reference to first embodiment in the flow process of Fig. 5, in the false colour element surveyed area set handling among the step S103 in the flow process of Fig. 2, location of pixels (x in step S304 at Fig. 5, when pixel value y) is indicated white saturated pixel, in step S305, false colour element surveyed area (x is set at white saturated pixel periphery
0, x
1, y
0, y
1), and carry out from false colour element surveyed area (x
0, x
1, y
0, y
1) detect the processing (the step S104 among Fig. 2 and the flow process among Fig. 7) of false colour element.
In the processing of aforesaid first embodiment, use default false color image surveyed area to determine that information processing is used to be arranged on the false color image surveyed area (x of white saturated pixel periphery
0, x
1, y
0, y
1), described default false color image surveyed area is determined information such as look-up table (LUT), wherein according to the feature of the optical system of image capture system false colour element surveyed area corresponding to image-region is set.
As mentioned above, the white saturated pixel that comprises in captured image is a high luminance pixel, that is, and and the saturated pixel in the image.For example, can be set at view data in the 0-255 scope, will be defined as white saturated pixel, and at white saturated pixel periphery false colour element surveyed area (x is set according to look-up table above all pixels of 250 for its brightness value
0, x
1, y
0, y
1).But, white saturated pixel, the high luminance pixel in the promptly captured image is not all to produce according to the object with same brightness (brightness), but the object with the lightness that is greater than or equal to certain lightness produces white saturated pixel.That is, to a certain degree senior its very high level part that is clipped to object lightness all is revealed as white saturated pixel in captured image.
But, wherein in zone that purple fringe appears in white saturated pixel periphery according to the change of the lightness level of object, promptly the difference in the lightness in the marginal portion changes.Even for the identical white saturated pixel in captured view data, have with the differential corresponding marginal portion of the lightness between white saturated pixel and neighborhood pixels thereof in the white saturated pixel of big luminosity equation cause big purple fringe to produce the zone, and the white saturated pixel that has little luminosity equation in the marginal portion causes little purple fringe to produce the zone.Therefore, generally, for the white saturated pixel with high object lightness level (actual lightness), purple fringe produces the zone and trends towards reducing, and for the white saturated pixel with very high object lightness level (actual lightness), produces the zone and trends towards increasing.
Therefore, even for the identical white saturated pixel in captured view data, change according to the actual lightness of object and false colour element surveyed area (x is set
0, x
1, y
0, y
1) can realize the reliable of false colour element and effectively detection.In a second embodiment, detect or estimate the actual lightness (intrinsic brilliance) of object, and false colour element surveyed area is set according to the intrinsic brilliance of object.
As mentioned above, have with the white saturated pixel of the big luminance difference of the differential corresponding marginal portion of the brightness between white saturated pixel and neighborhood pixels thereof and cause big purple fringe to produce the zone, and the white saturated pixel that has little luminance difference in the marginal portion causes little purple fringe to produce the zone.Therefore, need to consider that the luminance difference in the marginal portion produces scope so that determine purple fringe.On the principle, because purple fringe takes place the some expansion difference that causes owing to optical wavelength.The point expansion difference that causes owing to wavelength depends on the setting of type, focal length and the aperture etc. of lens, the position in image.Therefore, as fruit dot expansion because optical wavelength and slight modification even there then not have the marginal portion of too big luminosity equation also to produce therein in theory to be false colored, but does not produce the vacation colour that the people can perceive when some expansion difference is very little.But even when the some expansion difference that causes owing to wavelength is very little, if the edge brightness difference becomes saturated to hi-lite greatly, then little difference is exaggerated, and appreciable false colored in big region generating.Therefore, must consider luminance difference and lens type, setting and location of pixels, so that accurately definite zone that wherein can produce purple fringe in the marginal portion.
Present embodiment has overcome such problem, and the method that detects and proofread and correct the purple fringe that produces at the high luminance pixel periphery more accurately is provided.As among first embodiment, configuration shown in Figure 1 is applicable to the configuration of image processing equipment in the present embodiment.Described disposed of in its entirety sequence is also in the same manner as in the first embodiment as being performed according to the processing of flow chart shown in Figure 2.
The difference of second embodiment and first embodiment is the processing among the step S305 in the flow process shown in Fig. 5, and described processing is corresponding to the false colour element surveyed area set handling among the step S103 in flow process shown in Figure 2.That is, false colour element surveyed area (x is set
0, x
1, y
0, y
1) the processing difference.Other is handled, i.e. processing the step S103 in the flow process of the Fig. 2 in first embodiment is similar with the processing in first embodiment.
In the present embodiment, the false colour element surveyed area (x among the step S305 in the flow process of Fig. 5
0, x
1, y
0, y
1) set handling is performed as the processing that the best false colour element surveyed area corresponding with the lightness (intrinsic brilliance) of object is set.Below, the details of the processing that false colour element surveyed area is set in the present embodiment is described with reference to Figure 15 and accompanying drawing subsequently.
The principle of the generation of false colored (purple fringe) is described with reference to Figure 15.Figure 15 (a) and Figure 15 (b) are the lightness (intrinsic brilliance) of wherein vertical axis indicated object and the figure that trunnion axis is represented the location of pixels of captured view data.Figure 15 (a) comprises the lightness (intrinsic brilliance)=500 of object, and this part is revealed as saturated pixel, i.e. white saturated pixel in captured image.Saturation threshold line 510 presentation video capture devices in Figure 15 (a) and Figure 15 (b) are with respect to the saturation threshold of incident light quantity.That is, when more than or equal to the light quantity incident of saturation threshold line 510, described image is saturated and become white saturated image.
Figure 15 (a) shows the situation of wherein taking object as described below, described object has neighbouring part, the lightness of described neighbouring part is suddenlyd change between object lightness (intrinsic brilliance) 500 and lightness (intrinsic brilliance) 10, that is, its intrinsic brilliance of described object changes to 50 from 500.Figure 15 (b) shows and wherein takes its actual lightness from 5000 situations that change to 50 object.In Figure 15 (a), the part with intrinsic brilliance 500 be positioned at saturation threshold line 510 or on, and this part becomes white saturated part.In Figure 15 (b), the part with intrinsic brilliance 5000 be positioned at saturation threshold line 510 or on, and this part becomes white saturated pixel, they are expressed as the pixel with identical white saturated pixel in captured image.
In Figure 15 (a), solid line (G) is a green channel light quantity line (G) 511, indication is incident on the amount of the green channel light on the image-capture device, and dotted line (B) is a blue channel light quantity line (B) 512, and indication is incident on the amount of the blue channel light on the image-capture device.In Figure 15 (b), similarly, solid line (G) is a green channel light quantity line (G) 521, and indication is incident on the amount of the green channel light on the image-capture device, dotted line (B) is a blue channel light quantity line (B) 522, and indication is incident on the amount of the blue channel light on the image-capture device.
The state that Figure 15 (a) and Figure 15 (b) diagram wherein use same lens to take.Because use same lens to be used for described shooting, so the point of green channel and blue channel expansion profile (profile) is identical.But, the intrinsic brilliance of the intrinsic brilliance of the object among Figure 15 (b) and the object in Figure 15 (a) differs from one another, and it is much bigger to be incident on the light quantity that the light amount ratio on the image-capture device is incident on the image-capture device when the object among shooting Figure 15 (b) during the object in taking Figure 15 (b).
When incident light quantity was big, the remarkable rank that captured image is presented between green and the blueness was poor.The part that green in captured image and the rank difference between the blueness are big therein produces purple fringe.Therefore, in Figure 15 (a), about 10 pixel portion on the right side of white saturated pixel are the parts that wherein produce purple fringe probably, and in Figure 15 (b), about 20 pixel portion on the right side of white saturated pixel are the parts that wherein will produce purple fringe probably.Hi-lite in Figure 15 (a) and 15 (b) also has wherein the significant part that becomes of the rank difference between green and blueness, and this part is revealed as white saturated pixel in captured image, and does not become the purple fringe pixel.
By this way, cause through being everlasting to produce purple fringe around the white saturated hi-lite, and poor in the brightness of marginal portion depended in the zone that wherein produces purple fringe.Illustrate that with reference to Figure 16 brightness value and purple fringe in the captured image of the object with intrinsic brilliance corresponding with Figure 15 (a) and 15 (b) produce the zone.Figure 16 (a) and 16 (b) are the figure corresponding to Figure 15 (a) and Figure 15 (b).Figure 16 (a) changes to 50 object corresponding to its intrinsic brilliance from 500, and Figure 16 (b) changes to 50 object corresponding to its intrinsic brilliance from 5000.The brightness value of the view data that obtains when using same lens to take those two images distributes and becomes shown in Figure 16 (c).
In Figure 16 (c), vertical axis is indicated the brightness value (that is, the brightness value of captured image) of captured view data, trunnion axis indication location of pixels.In view data, intensity level 255 is intensity levels of saturation level, intrinsic brilliance 500 and intrinsic brilliance 5000 all cause on captured image the pixel with saturated intensity level of about 255, be white saturated pixel.Purple fringe appears at the part of the vicinity of white saturated pixel.This zone that can produce purple fringe is corresponding with about 10 pixel portion in the regional A shown in Figure 16 (c) of the object of the intrinsic brilliance 500 shown in Figure 16 (a), and corresponding with about 20 pixel portion in the area B shown in Figure 16 (c) of the object of the intrinsic brilliance 5000 shown in Figure 16 (b).
By this way, causing that white saturated hi-lite periphery produces purple fringe, and the luminance difference in the marginal portion is depended in the zone that wherein produces purple fringe.Need to determine unsaturated original brightness, so that determine luminance difference in the marginal portion.As a kind of method of this problem of processing, can use such method, wherein use the image of taking by image-capture device with muting sensitivity transducer.
On concrete, use the image-capture device of wherein having arranged a plurality of transducers with different sensitivity for pixel at same position.For example, two kinds of transducers have been arranged, i.e. normal sensibility transducer and muting sensitivity transducer.The image of taking by the normal sensibility transducer is used as captured image, and determines to have become the primary object brightness (intrinsic brilliance) of part of the white saturated pixel of the image of taking by the normal sensibility transducer according to the image of taking by the muting sensitivity transducer.Even for having the object that causes white saturated high brightness by the normal sensibility transducer, the image of taking by the muting sensitivity transducer is because muting sensitivity and do not cause saturated in vain in captured image, and is outputted as the pixel of the brightness value with the value of not reaching capacity.Therefore, for object with the intrinsic brilliance 500 shown in Figure 16 (a) and the object with the intrinsic brilliance 5000 shown in Figure 16 (b), described image can be outputted as the view data with different separately photographic images brightness values.
The processing of estimating the intrinsic brilliance of object according to the image that uses the muting sensitivity transducer to take is described with reference to Figure 17.Figure 17 (a) and Figure 17 (b) are the figure corresponding to Figure 15 (a) and 15 (b) and Figure 16 (a) and 16 (b).Figure 17 (a) changes to 50 object corresponding to its intrinsic brilliance from 500, and Figure 17 (b) changes to 50 object corresponding to its intrinsic brilliance from 5000.The brightness value of the view data that obtains when using the muting sensitivity transducer to take those two images distributes and becomes shown in Figure 17 (p) and 17 (q).The brightness value of the view data that obtains when taking those two images when using the normal sensibility transducer distributes and becomes shown in Figure 17 (r).Figure 17 (r) is the figure that is similar to Figure 16 (c).For any situation of intrinsic brilliance 500 and 5000, captured image brightness value is output at roughly 255, promptly is outputted as white saturated pixel.
But, shown in Figure 17 (p) and 17 (q), in the brightness value of the view data that obtains when using the muting sensitivity transducer to take distributes, on the image that normal sensibility is taken, cause that white saturated image section does not become white saturated pixel, and be outputted as the pixel that has less than the brightness value of saturated pixel value.In the graphic example of institute, in as the data shown in Figure 17 (p) of view data of the object with the intrinsic brilliance 500 shown in Figure 17 (a), image section with intrinsic brilliance 500 is outputted as the view data with captured image brightness 5, and in as the data shown in Figure 17 (q) of view data of the object with the intrinsic brilliance 5000 shown in Figure 17 (b), the image section with intrinsic brilliance 5000 is outputted as the view data with captured image brightness value 50.By this way, when using the muting sensitivity transducer to take, even it is white saturated to cause that by normal transducer white saturated part does not cause yet because of a large amount of incident light, and is outputted as the pixel that has less than the class value of saturated brightness value.Therefore, the view data of to a certain degree bright object and very bright object can be outputted as the view data with different brightness values.In the present embodiment, shot image data is used as benchmark image under the condition different with the condition of normal photographing process, to determine and to estimate the object intrinsic brilliance of the white saturated pixel part in the image of normal photographing, and according to estimated result, setting can produce the zone of false colored (purple fringe).
The example of the benchmark image of the intrinsic brilliance of wherein using the muting sensitivity transducer to obtain to be used for to determine object has been described.As another example, described configuration can be so that carry out shooting to obtain benchmark image with dark exposure by improving shutter speed.Even white when saturated when occurring in the image in normal photographing, promptly in the image of taking with correct exposure, occur whitely when saturated, the shooting results that uses dark exposure is the pixel value of white saturated and output less than saturated brightness value not occur.Therefore, the image of taking by the shutter speed that improves also can be used as benchmark image.For example, when with the time for exposure in the correct exposure about 1/30 and the image taken when being used as benchmark image can be estimated the object intrinsic brilliance of pixel saturated in correct exposure.
As mentioned above, the image of under low conditions of exposure, taking, such as image of taking by the muting sensitivity transducer or the image taken by high shutter speed, be used as benchmark image, estimate the intrinsic brilliance of object according to the benchmark image brightness value of the part corresponding, and the pixel detection zone of best false colored (purple fringe) is set according to estimated intrinsic brilliance with the white saturated pixel in the view data of normal photographing.
Figure 18 shows in the object intrinsic brilliance true brightness of saturated pixel (white) and can produce an example of the relation between the scope (apart from the distance of saturated pixel in vain) of false colored (purple fringe).Figure shown in Figure 180 is corresponding to example shown in Figure 17, and showing the scope (apart from the distance of white saturated pixel) that can produce false colored (purple fringe) when the object intrinsic brilliance is 500 is white saturated pixel 10 pixels of distance, and the scope (apart from the distance of white saturated pixel) that can produce false colored (purple fringe) when the object intrinsic brilliance is 5000 is distance saturated pixel 20 pixels in vain.In the example shown in the described figure is an example, and according to the feature of the optical system of image capture system, is used for the data of measuring in advance in the object intrinsic brilliance and can produces relation between the scope of false colour (purple fringe).
The process of false in the present embodiment colour element surveyed area set handling is described with reference to the flow process shown in Figure 19.Flow process shown in Figure 19 is the part for the flow process among the graphic Fig. 5 of aforesaid first embodiment.That is, the flow process among Figure 19 shows step S304 in Fig. 5 and the processing among the S305, determines whether that promptly institute's concerned pixel is white saturated pixel and the false colour element surveyed area (x that is arranged on described pixel periphery when it is white saturated pixel
0, x
1, y
0, y
1) processing.In a second embodiment, the processing of this part is performed as the processing different with first embodiment.
The processing sequence of second embodiment will be described according to the flow process among Figure 19.Step S304 is the processing that is similar to first embodiment, and determines the pixel value at the white saturated shielding S of set location of pixels.For white saturated shielding S (x, y)=(x y), handles and proceeds to step S314 (see figure 5) to carry out the processing for next pixel false set location of pixels.If (x is true y) to S, then handles proceeding to step S305a.(x, y) (x is a fictitious time y), this means location of pixels (x, the non-white saturated pixel of pixel value indication y) for location of pixels as white saturated shielding S.(x, y) (x is a true time y), and (x, pixel value y) is indicated white saturated pixel to this means location of pixels for location of pixels as white saturated shielding S.
(x, y) (x y) is true time, at step S305a, obtains corresponding to white saturated pixel position (x, the pixel value of benchmark image pixel y) for location of pixels as white saturated shielding S.Benchmark image for example refers to image of taking by aforesaid muting sensitivity transducer or the image of taking under low conditions of exposure, the described image of taking under low conditions of exposure is for example with high-speed shutter speed and with correct exposure or be lower than correct exposure and the image taken.According to benchmark image, obtain and the pixel (x, y) pixel value of Dui Ying benchmark image pixel that in as normal photographing image, are confirmed as white saturated pixel processed image.
Among step S305bs, according to the pixel value of benchmark image pixel, use look-up table (LUT) determine false colour element surveyed area (x thereafter,
0, x
1, y
0, y
1).Figure 20 shows the example of the data structure of the look-up table (LUT) that uses in the present invention.Look-up table shown in Figure 20 (LUT) has tableau format, and wherein, connection is relative to each other for " brightness of benchmark image ", " intrinsic brilliance of object " and " can produce false colored (PF: pixel coverage purple fringe) ".This form is shown as corresponding to top table with reference to the described example of Figure 15-17.
For example, illustrate when the brightness in benchmark image is " 5 ", the object intrinsic brilliance is " 500 ", and in this case " can produce false colored (PF: pixel coverage purple fringe) " is in the scope of white 10 pixels of saturated pixel of distance.Similarly, illustrate when the brightness in benchmark image is " 50 ", the object intrinsic brilliance is " 5000 ", and in this case " can produce false colored (PF: pixel coverage purple fringe) " is in the scope of white 20 pixels of saturated pixel of distance.
In the step S305b shown in Figure 19, for example, determine " can produce false colored (PF: pixel coverage purple fringe) " to be set from the form shown in Figure 20 according to the brightness benchmark image, and according to the result, with false colour element surveyed area (x
0, x
1, y
0, y
1) be defined as comprising and can produce false colored (PF: the zone of pixel coverage purple fringe).
Though form shown in Figure 20 has the structure of the intrinsic brilliance data that comprise object, these data are unessential.If described form be configured to allow according to with processed image, i.e. the image of normal photographing, on the brightness of benchmark pixel of white saturated pixel correspondence, and definite false colour element surveyed area (x
0, x
1, y
0, y
1), get final product.That is, described form can be configured to look-up table (LUT), wherein, " brightness of benchmark image " and " can produce false colored (PF: pixel coverage purple fringe) " or false colour element surveyed area (x
0, x
1, y
0, y
1) connection is relative to each other.
The functional configuration of the digital signal processor (DSP) (corresponding to DSP shown in Figure 1 106) in the image processing equipment of the present embodiment of carrying out above-mentioned processing is described with reference to Figure 21 and 22.
Figure 21 is with first embodiment, with reference to the corresponding figure of the configuration of the described DSP of Figure 14, and Figure 22 shows the figure that the configuration of unit 403 is set at the false colour element surveyed area shown in Figure 21.As shown in figure 21, as among first embodiment, digital signal processor in the present embodiment (DSP) has color conversion cell 401, saturation detection unit 402, false colour element surveyed area are provided with unit 403, false colored (purple fringe) detecting unit 404, pixel value correcting unit 410 and color inverse conversion unit 405 in vain.Pixel value correcting unit 410 comprises false colored (purple fringe) compensated interpolation unit 411 and false colored (purple fringe) Fuzzy Processing unit 412.
Digital signal processor (DSP) 106 receives view data (RGB from input picture frame memory 420
In), and the color space of the RGB image of color conversion process unit 401 conversion inputs, and be separated into luminance component and color component.The processing of being carried out by color conversion process unit 401 is corresponding to the processing in step S101 in the flow chart shown in Fig. 2.Input picture RGB
InBe converted, and be separated into luminance component image L
InWith the color component image C
InThat is, the RGB coloured image of input is converted into color space with luminance component and the color space with color component.Can use YCbCr or CIE L
*a
*b
*Wait and be used for described color space.
Then, according to by color conversion process unit 401 data converted, white saturation detection unit 402 detects white saturated pixel portion takes place.This is handled corresponding to the step S102 in flow chart shown in Figure 2.On concrete, as described in reference to flow chart shown in Figure 4, detect and have the pixel of the brightness that is greater than or equal to the predetermined luminance threshold value, and produce the white saturated shielding S that is used to discern white saturated pixel (x, y).
False colour element surveyed area is provided with unit 403 and carries out the processing that false colour element surveyed area is set in the part of the white saturated pixel periphery that is detected by white saturation detection unit 402.This regional set handling is to determine wherein can satisfy white saturated shielding S (x, y)=genuine white saturated pixel (x, y) periphery produces the processing in the zone of false colour, and as mentioned above, carries out the best false colour element surveyed area set handling of using benchmark image.
Illustrate that with reference to Figure 22 false colour element surveyed area is provided with the functional configuration of unit 403.As shown in figure 22, false colour element surveyed area is provided with unit 403 and has benchmark image analytic unit 711, look-up table stores unit 712 and false colour element surveyed area and determine processing unit 713.Benchmark image analytic unit 711 402 obtains with the white saturated pixel position the processed image from white saturation detection unit, and obtains the brightness value of the benchmark image pixel corresponding with saturated pixel position in vain.Described benchmark image is from the image of taking with object that processed image is identical, but is the image taken with the muting sensitivity transducer or to be lower than the exposure of correct exposure, for example with the setting of high-speed shutter speed, and the image of shooting.
False colour element surveyed area determines that processing unit 713 receives and brightness value in the benchmark image pixel that the white saturated pixel position the processed image is corresponding from benchmark image analytic unit 711.According to the value that is received, false colour element surveyed area determines that processing unit 713 is by determining false colour element surveyed area with reference to the look-up table (LUT) of storage in look-up table stores unit 712.For example, as above described with reference to Figure 20, described look-up table (LUT) is such look-up table (LUT), wherein, and " brightness of benchmark image " and " can produce false colored (PF: pixel coverage purple fringe) " or " false colour element surveyed area (x
0, x
1, y
0, y
1) " connection is relative to each other.
According to brightness value in the benchmark image pixel that the white saturated pixel position in the processed image is corresponding, false colour element surveyed area determines that processing unit 713 determines false colour element surveyed area (x
0, x
1, y
0, y
1).False colour element surveyed area determines that processing unit 713 will (x, y) scope of Zhou Bian false colour element surveyed area be defined as four scalar value (x in white saturated pixel
0, x
1, y
0, y
1), and produce by remove the false colored region shielding P (seeing the handling process Fig. 5 and 19) that white saturated pixel obtains from that zone.This is handled and makes and can under the situation of considering practical object brightness (intrinsic brilliance) false colour element surveyed area be set, and realizes that thus best zone is provided with.Return referring to Figure 21, will continue the explanation of present embodiment.
False colored (purple fringe) detecting unit 404 uses by false colour element surveyed area the false colored region shielding P that unit 403 is provided with is set, and carries out and detect by each pixel being carried out the processing that color determines to handle the pixel that is confirmed as having false colored (purple fringe).That is, upgrade false colored region shielding P to produce the false colored region shielding P that can only discern the false colour element that to proofread and correct.This is handled corresponding to the step S104 in the flow chart shown in Fig. 2, and is performed according to the handling process shown in Fig. 7.As mentioned above, be confirmed as false colored color is arbitrarily, the purple that therefore wherein only has a specific color value be set to false colored configuration or wherein the configuration that is set to false colour of the multicolour such as green and purple be fine.
False colour (purple fringe) compensated interpolation unit 411 is carried out and is proofreaied and correct the processing that is defined as having the colored pixel of vacation by false colored (purple fringe) detecting unit 404.Described processing is corresponding to the step S105 in the flow chart shown in Fig. 2, and according to being performed in the handling process shown in Figure 11 and 12.This pixel value treatment for correcting comprises: according to the value of the pixel except the false colored and white saturated pixel around it compensated interpolation that pixel that is confirmed as having false colored (purple fringe) is carried out is handled (seeing Figure 10 and 11); Reduce the compensated interpolation that does not also have scheduled number of repetition and handle the saturated processing (seeing Figure 12) of color of false colored (purple fringe) pixel of proofreading and correct.
The processing that 412 pairs of false colored (purple fringe) Fuzzy Processing unit are undertaken by false colored (purple fringe) compensated interpolation unit 411 and data of proofreading and correct are carried out Fuzzy Processing.This is handled corresponding to the step S106 in the flow chart shown in Fig. 2, and according to being performed in the handling process shown in Figure 13.
False colored (purple fringe) Fuzzy Processing unit 412 extracts by false colored (purple fringe) detecting unit 404 and is defined as having false colored pixel, and applies blur filter to color component.For example, pixel (the x that false colored (purple fringe) Fuzzy Processing unit 412 is determined to comprise with processed, and carry out this mean value and be set to y) and the mean value of 9 pixels of 8 pixels on every side, with processed pixel (x, the rolling average filtration treatment of renewal pixel value y).
The luminance component image L of color inverse conversion unit 405 conversion original images
InWith color component image C, and export the image RGB that as a result of produces as above-mentioned result
OutBe used as the RGB coloured image.
When using present embodiment, can be for the suitable correction of false colored execution such as purple fringe that causes by chromatic aberation, wherein said chromatic aberation takes place for the image of being taken by video camera, and can realize producing and the outputting high quality view data.
In addition, according to second embodiment, estimate and the white corresponding object intrinsic brilliance of saturated pixel according to benchmark image, false colored (purple fringe) surveyed area corresponding to the object intrinsic brilliance is set, and detect and proofread and correct false colored (purple fringe) pixel in set zone.Therefore, can be provided with and regional corresponding false colored (purple fringe) surveyed area that wherein might produce false colored (purple fringe), and not make set false colored (purple fringe) surveyed area become too big or too little.Can be reliably and detect false colored (purple fringe) pixel effectively, improve the accuracy of image rectification, and raise the efficiency.
Describe the present invention in detail with reference to certain embodiments.But obviously those skilled in the art can make amendment to described embodiment in the scope that does not break away from essence of the present invention and replace.That is, the present invention is disclosed by way of example, therefore is not appreciated that determinate.Be to be understood that the scope of claim, so that understand essence of the present invention.
Can carry out series of processes described herein by hardware, software or its combining and configuring.When using software to carry out processing, can wherein write down the program of handling sequence by installing in the memory in the computer in being built in specialized hardware, or described program is installed to the all-purpose computer that can carry out various types of processing, carry out above-mentioned processing.
For example, can be on the storage medium such as hard disk or ROM (read-only memory) storing said program in advance.Perhaps, described program can be stored (record) on removable storage medium, such as floppy disk, CD-ROM (compact disk read-only memory), MO (magneto-optic) dish, DVD (digital versatile disc), disk or semiconductor memory temporarily or for good and all.Removable storage medium like this can be used as so-called " canned software " and is provided.
Except installing the described program to computer from above-mentioned removable storage medium, can also transmit described program to computer radio from download site, perhaps can transmit described program via wired connection to computer, so that the described program of Chuan Songing is received and is stored in by computer on the storage medium such as built-in hard disk by this way by network such as LAN (local area network (LAN)) and/or internet.
Various types of processing described herein not only comprises the processing of carrying out with time sequencing according to described sequence, and comprises according to the throughput of carrying out the equipment of handling or the processing of execution as required and at the same time or separately.Term " system " at this refers to the logical combination of a plurality of equipment, and is not limited to wherein comprise in same shell the system of each equipment.
Industrial applicibility
As mentioned above, according to configuration of the present invention, detect white saturated pixel from view data, and in institute The white saturated pixel periphery that detects arranges false colour element surveyed area, detects from set zone to have With the pixel such as color corresponding to the false colour of purple fringe, the pixel that detects is defined as false colored Pixel, and for the correction processing of the false colour element execution of identifying based on the value of surrounding pixel. Cause This can detect the contiguous false colored district such as purple fringe that produces in white saturated pixel effectively Territory, and the value of correction pixels partly. Can produce with the outputting high quality view data does not affect yet Whole image.
According to configuration of the present invention, can be to the false colour such as purple fringe that is caused by chromatic aberation Carry out suitable correction, wherein said chromatic aberation takes place for the image of being taken by video camera, and can Realize producing and the outputting high quality view data. Use traditional video camera, might shoot not nature Image, unless producing the setting of adjusting in the situation of purple fringe such as lens stop. But, When use when of the present invention, can effectively extract and proofread and correct in captured image, comprise such as purple The false colour of striped and so on, and can produce and the outputting high quality image. Therefore, needn't pay close attention to lens Purple fringe does not appear in aperture and focal length during taking, therefore so that can come with the higher free degree Carry out and take.
In addition, according to configuration of the present invention, it is corresponding with white saturated pixel right to estimate according to benchmark image Resemble intrinsic brilliance, false colored (purple fringe) surveyed area corresponding to the object intrinsic brilliance is set, and inspection Survey and proofread and correct false colored (purple fringe) pixel in set zone. Therefore, can arrange and wherein have May produce false colored (purple fringe) surveyed area corresponding to zone of false colored (purple fringe), and not make Get set false colored (purple fringe) surveyed area and become too big or too little. Can be reliably and effectively Detect false colored (purple fringe) pixel, improve the accuracy of image rectification, and raise the efficiency.
Claims (26)
1. image processing equipment comprises:
White saturation detection unit is used for detecting white saturated pixel from view data;
False colour element surveyed area is provided with the unit, is used at the white saturated pixel periphery that is detected by white saturation detection unit false colour element surveyed area being set;
False colored detecting unit is used for the zone that the unit is provided with is being set by false colour element surveyed area, and the pixel that will have the color corresponding with false colour is identified as false colour element; And
The pixel value correcting unit is used for the value according to surrounding pixel, and the false colour element that is detected by the colored detecting unit of vacation is carried out the pixel value treatment for correcting.
2. according to the image processing equipment of claim 1, wherein, described pixel value correcting unit has: the compensated interpolation unit is used for false colour element being carried out compensated interpolation handling according to pixel value on every side; And color Fuzzy Processing unit, be used for false colour element is carried out the color Fuzzy Processing.
3. according to the image processing equipment of claim 2, wherein, described pixel value correcting unit is selected the pixel except false colour element and white saturated pixel, and is carried out compensated interpolation according to selected pixel and handle from the pixel that exists at false colour element periphery.
4. according to the image processing equipment of claim 1, wherein, the colored detecting unit of described vacation is carried out the processing that the pixel that will have default specific color in the zone that the unit setting is set by false colour element surveyed area is identified as false colour element.
5. according to the image processing equipment of claim 1, wherein
Described false colour is a purple fringe, and
The colored detecting unit of described vacation is carried out the processing that the purple pixel in the zone that the unit setting is set by false colour element surveyed area is identified as false colour element.
6. according to the image processing equipment of claim 1,
Wherein, described false colour element surveyed area is provided with the unit, according to take with the aperture during the processed view data, focus information and from light center to white saturated pixel (x, among the data of distance y) at least one carried out the processing of determining at the false colour element surveyed area of the white saturated pixel periphery setting that is detected by white saturation detection unit.
7. according to the image processing equipment of claim 1,
Wherein, the processing of the white saturated pixel of selecting to have the brightness that is greater than or equal to predetermined threshold is carried out in described white saturation detection unit.
8. according to the image processing equipment of claim 1,
Also comprise the color conversion cell that is used to carry out data conversion treatment, described data conversion treatment is separated into luminance component view data and color component view data with input image data, and
Described image processing equipment is also carried out the pixel value treatment for correcting according to the translation data that is produced by described color conversion cell.
9. according to the image processing equipment of claim 1,
Wherein, described false colour element surveyed area is provided with the unit and has:
The benchmark image analytic unit, be used to receive benchmark image, and obtain brightness value with the benchmark image pixel that the white saturated pixel in the processed image is corresponding, in described benchmark image, do not have saturated brightness value with the pixel of position that the white saturated pixel in the processed image is corresponding, described benchmark image is and the photographic images that processed image object is identical; And
False colour element surveyed area determining unit is used for according to the brightness value of the benchmark image pixel of correspondence false colour element surveyed area being set, and described brightness value is obtained by the benchmark image analytic unit.
10. according to the image processing equipment of claim 9, wherein, described false colour element surveyed area determining unit is provided with false colour element surveyed area by using look-up table, in described look-up table, the brightness value of benchmark image and can produce false colored pixel coverage or the false colour element surveyed area connection that is relative to each other.
11. according to the image processing equipment of claim 9,
Wherein, described false colour element surveyed area determining unit execution is provided with the processing of wideer false colour element surveyed area along with the increase of the brightness value of the respective pixel in benchmark image.
12. according to the image processing equipment of claim 9,
Wherein, be used as benchmark image by using by the captured image of the image-capture device with muting sensitivity transducer or being lower than the image of taking under the low conditions of exposure of correct exposure, false colour element surveyed area determining unit obtains and benchmark image brightness value in the part that the white saturated pixel in the processed image is corresponding.
13. an image processing method comprises:
White saturation detection step detects white saturated pixel from view data;
False colour element surveyed area is provided with step, is arranged on the false colour element surveyed area of the white saturated pixel periphery that detects in the white saturation detection step;
The false colored step that detects is provided with in the zone that is provided with in the step at false colour element surveyed area, and the pixel that will have the color corresponding with false colour is identified as false colour element; And
The pixel value aligning step is according to the value of surrounding pixel, to carrying out the pixel value treatment for correcting in the false colored false colour element that detects the step detection.
14. according to the image processing method of claim 13, wherein, described pixel value aligning step comprises:
The compensated interpolation step is carried out compensated interpolation to false colour element and is handled according to pixel value on every side; And
Color Fuzzy Processing step is carried out the color Fuzzy Processing to false colour element.
15. image processing method according to claim 14, wherein, described pixel value aligning step is such step: select the pixel except false colour element and white saturated pixel from the pixel that exists at false colour element periphery, and carry out compensated interpolation according to selected pixel and handle.
16. image processing method according to claim 13, wherein, described false colored detection step is such step: carry out being provided with in the zone that is provided with in the step at false colour element surveyed area, the pixel that will have default specific color is identified as the processing of false colour element.
17. according to the image processing method of claim 13, wherein
Described false colour is a purple fringe, and
Described false colored to detect step be such step: carry out the processing that purple pixel in the zone that is provided with in the step is identified as false colour element is set at false colour element surveyed area.
18. according to the image processing method of claim 13,
Wherein, it is such step that described false colour element surveyed area is provided with step: according to take with the aperture during the processed view data, focus information and from light center to white saturated pixel (x, among the data of distance y) at least one carried out the processing of the false colour element surveyed area that the white saturated pixel periphery determine to detect is provided with in white saturation detection step.
19. according to the image processing method of claim 13,
Wherein, described white saturation detection step is such step: carry out the processing that selection has the white saturated pixel of the brightness that is greater than or equal to predetermined threshold.
20. the image processing method according to claim 13 also comprises
Carry out the color conversion step of data conversion treatment, described data conversion treatment is separated into luminance component view data and color component view data with input image data, and described image processing method also comprises according to carrying out the pixel value treatment for correcting by the translation data of color conversion cell generation.
21. according to the image processing method of claim 13,
Wherein, described false colour element surveyed area is provided with step and has:
The benchmark image analytical procedure, the input reference image, and obtain brightness value with the benchmark image pixel that the white saturated pixel in the processed image is corresponding, in described benchmark image, do not have saturated brightness value with the pixel of position that the white saturated pixel in the processed image is corresponding, described benchmark image is and the photographic images that processed image object is identical; And
False colour element surveyed area determining step is provided with false colour element surveyed area according to the brightness value of the benchmark image pixel of correspondence, and described brightness value obtains in the benchmark image analytical procedure.
22. image processing method according to claim 21, wherein, described false colour element surveyed area determining step is such step: by using look-up table false colour element surveyed area is set, in described look-up table, the brightness value of benchmark image and can produce false colored pixel coverage or the false colour element surveyed area connection that is relative to each other.
23. according to the image processing method of claim 21,
Wherein, described false colour element surveyed area determining step is such step: execution is provided with the processing of wideer false colour element surveyed area along with the increase of the brightness value of the respective pixel in benchmark image.
24. according to the image processing method of claim 21,
Wherein, in described false colour element surveyed area determining step, use is used as benchmark image by the captured image of the image-capture device with muting sensitivity transducer or being lower than the image of taking under the low conditions of exposure of correct exposure, obtains and benchmark image brightness value in the part that the white saturated pixel in the processed image is corresponding.
25. a computer program is used to make the computer carries out image processing, described program comprises:
White saturation detection step detects white saturated pixel from view data;
False colour element surveyed area is provided with step, is arranged on the false colour element surveyed area of the white saturated pixel periphery that detects in the white saturation detection step;
The false colored step that detects is provided with in the zone that is provided with in the step at false colour element surveyed area, and the pixel that will have the color corresponding with false colour is identified as false colour element; And
The pixel value aligning step is according to the value of surrounding pixel, to carrying out the pixel value treatment for correcting in the false colored false colour element that detects in the step that detects.
26. according to the computer program of claim 25,
Wherein, described false colour element surveyed area is provided with step and comprises:
The benchmark image analytical procedure, the input reference image, and obtain brightness value with the benchmark image pixel that the white saturated pixel in the processed image is corresponding, in described benchmark image, do not have saturated brightness value with the pixel of position that the white saturated pixel in the processed image is corresponding, described benchmark image is and the photographic images that processed image object is identical; And
False colour element surveyed area determining step is provided with false colour element surveyed area according to the brightness value of the benchmark image pixel of correspondence, and described brightness value obtains in the benchmark image analytical procedure.
Applications Claiming Priority (3)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP2004157098 | 2004-05-27 | ||
| JP157098/2004 | 2004-05-27 | ||
| JP303795/2004 | 2004-10-19 |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1806449A true CN1806449A (en) | 2006-07-19 |
| CN100576924C CN100576924C (en) | 2009-12-30 |
Family
ID=36867544
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN200580000559A Expired - Fee Related CN100576924C (en) | 2004-05-27 | 2005-04-25 | Image processing equipment and image processing method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN100576924C (en) |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102474626A (en) * | 2009-07-21 | 2012-05-23 | 佳能株式会社 | Image processing device for correcting chromatic aberration, method for processing image, program, and storage medium |
| CN102118621B (en) * | 2009-12-30 | 2012-11-21 | 华晶科技股份有限公司 | Method for eliminating color dislocation for digital image |
| CN103957392A (en) * | 2009-07-21 | 2014-07-30 | 佳能株式会社 | Image processing device and image processing method |
| CN104093011A (en) * | 2014-07-22 | 2014-10-08 | 浙江宇视科技有限公司 | Pseudo-color inhibition method and device |
| CN104247409A (en) * | 2012-04-24 | 2014-12-24 | 索尼公司 | Image processing device, image processing method, and program |
| CN104378612A (en) * | 2013-08-12 | 2015-02-25 | 联想(北京)有限公司 | Method for adjusting color balance and electronic device |
| JP2015088922A (en) * | 2013-10-30 | 2015-05-07 | リコーイメージング株式会社 | Image processing device, imaging device, image processing method, and image processing program |
| CN111199524A (en) * | 2019-12-26 | 2020-05-26 | 浙江大学 | Purple edge correction method for image of adjustable aperture optical system |
| CN115215709A (en) * | 2022-07-14 | 2022-10-21 | 攀枝花恒威化工有限责任公司 | Emulsion explosive waste treatment system and method |
| CN115274093A (en) * | 2022-07-26 | 2022-11-01 | 华东师范大学 | Method and system for generating reference pathology data set containing automatic labeling file |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102474628B (en) * | 2010-03-25 | 2014-09-17 | 富士胶片株式会社 | Image processing method and device |
-
2005
- 2005-04-25 CN CN200580000559A patent/CN100576924C/en not_active Expired - Fee Related
Cited By (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103957392B (en) * | 2009-07-21 | 2016-06-01 | 佳能株式会社 | Image processing equipment and image processing method |
| US8659673B2 (en) | 2009-07-21 | 2014-02-25 | Canon Kabushiki Kaisha | Image processing apparatus, image processing method, program, and storage medium for correcting chromatic aberration |
| CN103957392A (en) * | 2009-07-21 | 2014-07-30 | 佳能株式会社 | Image processing device and image processing method |
| CN102474626B (en) * | 2009-07-21 | 2014-08-20 | 佳能株式会社 | Image processing device for correcting chromatic aberration, method for processing image, program, and storage medium |
| CN102474626A (en) * | 2009-07-21 | 2012-05-23 | 佳能株式会社 | Image processing device for correcting chromatic aberration, method for processing image, program, and storage medium |
| CN102118621B (en) * | 2009-12-30 | 2012-11-21 | 华晶科技股份有限公司 | Method for eliminating color dislocation for digital image |
| US9542759B2 (en) | 2012-04-24 | 2017-01-10 | Sony Corporation | Image processing device, method of processing image, and image processing program including false color correction |
| CN104247409A (en) * | 2012-04-24 | 2014-12-24 | 索尼公司 | Image processing device, image processing method, and program |
| CN104247409B (en) * | 2012-04-24 | 2016-10-12 | 索尼公司 | Image processing apparatus, image processing method and program |
| CN104378612A (en) * | 2013-08-12 | 2015-02-25 | 联想(北京)有限公司 | Method for adjusting color balance and electronic device |
| CN104378612B (en) * | 2013-08-12 | 2017-09-29 | 联想(北京)有限公司 | A kind of method and electronic equipment for adjusting color balance |
| JP2015088922A (en) * | 2013-10-30 | 2015-05-07 | リコーイメージング株式会社 | Image processing device, imaging device, image processing method, and image processing program |
| CN104093011B (en) * | 2014-07-22 | 2016-09-14 | 浙江宇视科技有限公司 | A kind of pseudo color inhibiting method and device |
| CN104093011A (en) * | 2014-07-22 | 2014-10-08 | 浙江宇视科技有限公司 | Pseudo-color inhibition method and device |
| CN111199524A (en) * | 2019-12-26 | 2020-05-26 | 浙江大学 | Purple edge correction method for image of adjustable aperture optical system |
| CN111199524B (en) * | 2019-12-26 | 2023-03-17 | 浙江大学 | Purple edge correction method for image of adjustable aperture optical system |
| CN115215709A (en) * | 2022-07-14 | 2022-10-21 | 攀枝花恒威化工有限责任公司 | Emulsion explosive waste treatment system and method |
| CN115274093A (en) * | 2022-07-26 | 2022-11-01 | 华东师范大学 | Method and system for generating reference pathology data set containing automatic labeling file |
| CN115274093B (en) * | 2022-07-26 | 2023-06-23 | 华东师范大学 | Method and system for generating reference pathology data set containing automatic labeling file |
Also Published As
| Publication number | Publication date |
|---|---|
| CN100576924C (en) | 2009-12-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN1806449A (en) | Image processing apparatus, image processing method, and computer program | |
| RU2338330C2 (en) | Device for image processing, method for image processing and computer program | |
| CN1689340A (en) | Digital still camera and image correction method | |
| JP4770154B2 (en) | Image processing apparatus, image processing method, and computer program | |
| CN1184796C (en) | Image processing method and equipment, image processing system and storage medium | |
| CN1207895C (en) | Image processing method, apparatus and recording medium for recording & executing the same method program | |
| CN1595958A (en) | Image quality correction apparatus and image quality correction method | |
| CN1114888C (en) | Image processing method and device, image processing system,and memory medium | |
| CN1102028A (en) | Object tracking apparatus | |
| CN1857008A (en) | Image processing method | |
| CN1662071A (en) | Image data processing in color spaces | |
| CN1643936A (en) | Image processing device, image processing method, and image processing device manufacturing method | |
| CN1744675A (en) | Image input device | |
| CN101032159A (en) | Image processing device, method, and image processing program | |
| CN1526117A (en) | Method and system for producing formatted data related to defect of appliances of a set of appliances | |
| CN1890987A (en) | Projector color correcting method | |
| CN1885952A (en) | Image sensing apparatus and image processing method | |
| CN1871847A (en) | Signal processing system, signal processing method, and signal processing program | |
| CN1732682A (en) | Image processing device and image processing program | |
| CN101076085A (en) | Method and apparatus for image capturing and electronic apparatus using the same | |
| CN1770829A (en) | Signal processing apparatus and signal processing method for solid-state image pickup element and image pickup apparatus | |
| CN1874418A (en) | Image processing apparatus, image processing method, and image processing program | |
| CN1812481A (en) | Image processing apparatus and its method | |
| CN1314266C (en) | Image pickup apparatus | |
| CN1993707A (en) | Image processing method and apparatus, image sensing apparatus, and program |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20091230 Termination date: 20190425 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |