JP2009171585A - Color recovery method and system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a color recovery method and system that restores the colors of digital video images distorted due to lighting into colors under standard lighting, thereby overcoming the specific color problem that occurs in digital video images. <P>SOLUTION: The present invention comprises the steps of (a) determining criterion regions, taking into consideration the variance values for each color channel of video images to be restored; (b) computing relative surface reflectance for the determined criterion regions and computing the adjustment coefficients for excluding the distortion effects of a light source, by using the computed relative surface reflectance; and (c) recovering the colors for the video images which are to be restored by using the adjustment coefficient. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a color restoration method and system. More particularly, the present invention relates to a color restoration method and system for restoring the color of a digital video distorted by illumination using the principle of realizing the color constancy of a human visual system.

  In general, when image distortion is caused by illumination, it is theoretically impossible for any system to accurately restore the color value of the image regardless of the distortion. The reason is that the color value is acquired under stable standard lighting conditions, or the color value is defined from the surface reflectance of the spectrum, and the image at this time does not include the light source itself, but the surface of the object. This is because the light information reflected from the light is included. However, according to the known color constancy function, the human visual system can reduce or minimize the effect of lighting distortion to obtain a value that closely approximates the unique color value of the video. Can be recognized.

  Conventionally, many attempts have been made to preferably realize such a human visual system as a video processing technique, but satisfactory results have not yet been obtained. Most conventional digital video processing devices and digital video processing software apply very complex color correction techniques that are far from human vision systems. Color correction technology applies various signals obtained mainly from the visual and technical viewpoints, but sometimes considers manual operation by trial and error. However, the color correction technique has a problem that its application range is limited. The reason is that if the operating parameter has a value within the specified range, the conventional system operates normally, but if the operating parameter has a value that deviates from the specified range, the performance of the conventional system will be drastically reduced. This is because it gets worse. Sometimes we do manual work to solve such problems of the conventional system, but in this case, there is a problem that it takes a lot of work and requires a lot of investment in terms of time and cost .

  In order to solve such problems of the prior art, Korean Patent Registration Publication No. 442,320 discloses a color restoration method for digital video. According to the present invention, the color restoration method uses an intra channel-ratio scale, and uses an equi-max algorithm to solve the unique color problem. The unique color problem usually occurs when, for example, a particular color appears remarkably over an entire area or a partial area of a distorted digital image so that blue is conspicuous in a coastal scene. The unique color problem occurs when an object has a relatively unique uniform color in an enlarged photograph. However, the equi-max algorithm according to the invention does not work very well on a stable criterion base. Therefore, there is a limit to overcoming the inherent color problem that occurs when the digital image is distorted by multiple light sources or the bias to the digital image is large, and readjustment of the correction parameters where the inherent color problem appears on the surface is avoided. Absent.

  The present invention, which has been derived to solve the above-described problems, restores the color of a digital image distorted by the lighting effect to a color under standard illumination, and causes a unique color problem (unique color) that occurs in the digital image. It is an object of the present invention to provide a color restoration method and system for overcoming (problem).

  In order to achieve the above object, the color restoration method according to the present invention includes: a) determining a compliant area in consideration of a dispersion value for each color channel of a video to be restored; and b) for the determined compliant area. Calculating a relative surface reflectivity, and using the calculated relative surface reflectivity to calculate an adjustment factor for eliminating the effect of an external light source; and c) using the adjustment factor And restoring the color of the video to be restored.

  In addition, the color restoration system according to the present invention includes a compliant region determination unit that determines a region where the sum of dispersion values of each color channel of a video to be restored is maximized as a compliant region; An adjustment coefficient calculation unit for calculating an adjustment coefficient for performing the adjustment; and a video restoration unit for restoring the color of the video to be restored using the adjustment coefficient.

  According to the present invention, if a predetermined number or more of surface elements having other reflectances are provided in a predetermined area of a digital image, the digital due to the spectral bias of the light source that cannot be solved by the conventional color correction technique. The distorted color of the video can be easily and quickly compensated regardless of the inherent color problem. In addition, it is easy to restore the color of a digital image distorted by multiple light sources, and the inherent color problem that occurs when the bias to the digital image increases can be solved.

1 is a block diagram of a color restoration system according to a preferred embodiment of the present invention. 3 is a flowchart of a color restoration method according to a preferred embodiment of the present invention. FIG. 5 is a reference diagram for a process of restoring a unique color for each channel of a digital image according to an exemplary embodiment of the present invention. FIG. 5 is a reference diagram for a color restoration result of an image biased in red according to a preferred embodiment of the present invention. FIG. 6 is a reference diagram for a color restoration result of a green biased image according to an exemplary embodiment of the present invention. FIG. 6 is a reference diagram for a color restoration result of a blue-biased image according to an exemplary embodiment of the present invention.

  Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. First, when adding reference numerals to the components in each drawing, it should be noted that the same components are given the same reference numerals as much as possible when shown on other drawings. Further, in the description of the present invention, when it is determined that a specific description related to a known configuration or function is likely to obscure the gist of the present invention, detailed description thereof is omitted. Furthermore, although preferred embodiments of the present invention will be described below, it is needless to say that the technical idea of the present invention is not limited or limited thereto and can be variously implemented by those skilled in the art.

  FIG. 1 is a block diagram of a color restoration system according to a preferred embodiment of the present invention. The color restoration system 10 restores the color of a digital image distorted by illumination using the principle of realization of color constancy of the human visual system, and operates according to the following three basic assumptions.

  First, since the problem of color constancy cannot be solved by using an inverse function, it must be approached by a method for obtaining an approximate value rather than calculating an accurate value. The human visual system uses a light adaptation method to obtain an approximation. The color restoration system 10 according to the present invention considers the light adaptation method when restoring color of a distorted digital image.

  Secondly, the effect of the light source by illumination acts uniformly on all parts of a single image. Therefore, if the approximate color correction value in a specific area is known, the approximate color correction value in another area can be easily obtained.

  Thirdly, natural images mostly have a light reflection spectrum on the surface of the object. If the number of colors included in the video is less than or equal to a predetermined number, the system can determine whether the colors that can be recognized by the eyes are due to the inherent surface characteristics of the object (for example, surface reflectance) or distortion caused by the light source. Can not. In order to overcome such problems, the present invention assumes that a predetermined type of color is provided on the surface of an object in an image.

  As shown in FIG. 1, the color restoration system 10 includes a compliant region determination unit 20, a ratio scale calculation unit 30, a surface reflectance calculation unit 40, a color restoration unit 50, an adjustment coefficient calculation unit 60, a video restoration unit 70, A correction unit 73 and a second correction unit 75 are included.

The compliant region determination unit 20 determines a region where the sum of dispersion values for each color channel of the video input for color restoration is maximized as a compliant region.
The ratio scale calculation unit 30 includes an inter-channel ratio scale (IR) that is a ratio of a hue value for each color channel of each pixel constituting the reference area to a hue value for each color channel of the entire reference area. ). In the embodiment of the present invention, the ratio scale calculation unit 30 performs the function in consideration of the first assumption and the second assumption described above.

The surface reflectance calculator 40 receives the intra-channel ratio scale and calculates a relative surface reflectance that is an approximate value for the surface reflectance of each pixel in the reference region.
The color restoration unit 50 multiplies the relative surface reflectance and the light intensity of each pixel in the reference region to restore the unique surface color of each pixel in the reference region.

The adjustment coefficient calculation unit 60 calculates an adjustment coefficient for eliminating the effect of the light source by using the restored unique surface color for each color channel of the reference region and the hue value for each color channel of the reference region.
The video restoration unit 70 restores the video input for color restoration by using the hue value of each pixel constituting the video inputted for color restoration and the adjustment coefficient.

The first correction unit 73 performs overflow control (Overflow control) when the hue value for each pixel of the restored image exceeds a limit value specified in 0 to 255, and for each pixel of the restored image. Correct the hue value of.
The second correction unit 75 re-corrects the corrected hue value for each pixel in consideration of the corrected hue value for each pixel and the output modulation factor (gain factor).

  FIG. 2 is a flowchart of a color restoration method according to a preferred embodiment of the present invention. As shown in FIG. 2, the color restoration method includes the following steps performed in time series in the color restoration system 10.

In S10, the compliant area determination unit 10 determines a compliant area for the input video for color restoration. The compliant area is an area where the sum of variance values for each color channel is the maximum in the input video. Based on statistical properties, this compliant area is likely to have a more diverse color than other areas of the video. Therefore, if the compliant area is used, it is possible to minimize the lighting effect caused by the main color expanded to a relatively large area in the process of calculating the color adjustment coefficient, thereby effectively solving the inherent color problem. be able to.
The calculation of the reference area is performed by the following mathematical formula 1.

Here, σ ² is the sum of the dispersion values of the three channels, σ ² _R is the dispersion value of the R channel, σ ² _G is the dispersion value of the G channel, and σ ² _B is the dispersion value of the B channel.
The method for determining the reference area is as follows.

First, the video to be restored is divided into a plurality of sub-videos. Also, a sum of dispersion values of each color channel for each of the plurality of divided sub-pictures is calculated, and a sub-picture that maximizes the sum of the dispersion values is determined as a first sub-picture.
Next, the first sub video is subdivided into a plurality of sub videos. In addition, a sum of dispersion values of each color channel for each of the subdivided sub-pictures is calculated, and a sub-picture having the maximum sum of the dispersion values is determined as a second sub-picture.
Finally, the first sub-video and the second sub-video are compared, and the first sub-video is determined as a reference area when the sum of the variance values of the first sub-video is larger.

  If the sum of the variance values of the second sub-video is larger, the second sub-video is subdivided into a plurality of sub-videos, and then the variance value of each color channel in the sub-divided sub-video is determined. The reference region is determined by repeating the division until the sum of the variance values in each color channel of the sub video with the maximum sum is smaller than or equal to the sum of the variance values of each color channel in the sub video before the re-division. be able to.

  Further, another method for determining the compliant region is to extract sub-images by applying windows of different sizes to the image to be restored, and to determine a variance value for each color channel for each of the extracted sub-images. After calculating the sum of the sub-images, the sub-video with the maximum sum of the dispersion values for each color channel is determined as the reference region.

In S <b> 20, the ratio scale calculation unit 30 calculates an intra-channel ratio scale (IR: Inter-channel Ratio scale). At this time, the in-channel ratio scale calculated by the ratio scale calculation unit 30 is derived from the hue value for each color channel of each pixel constituting the reference area and the ratio value between the hue values for each color channel of the reference area. The channel-specific hue value for each video pixel means a ratio scale that represents the relative intensity of the pixels in the respective color channel. Therefore, it can be said that the ratio value is the same as the normalized hue value of each pixel for each channel. In the embodiment of the present invention, it is preferable that the ratio scale calculation unit 30 considers the three-channel hue value of each pixel when calculating the in-channel ratio scale.
The calculation of the in-channel ratio scale is performed by the following mathematical formula 2.

ここで、Ｖ_Ｋは準拠領域の単一ピクセルのＫ色チャネルに対する色相値、

は準拠領域全体のＫ色チャネルの色相値、およびＩＲ_Ｋは準拠領域のピクセル別のＫ色チャネルに対するチャネル内比率尺度である。

Where V _K is the hue value for a single pixel K color channel in the compliant region,

Is the hue value of the K color channel for the entire compliant region, and IR _K is an in-channel ratio measure for the K color channel by pixel of the compliant region.

  The mathematical formula 2 is used to calculate an in-channel ratio scale which is a core calculation component in the color restoration method according to the present invention, and the hue value of each pixel constituting the reference region is converted into a ratio value by the mathematical formula 2. The

In S30, the surface reflectance calculation unit 40 calculates a relative surface reflectance for replacing the surface reflectance that determines the intrinsic color of the object surface.
The relative surface reflectance is calculated according to the following mathematical formula 3.

ここで、Ｐ_Ｋは準拠領域の各ピクセルのＫ色チャネルにおける相対的表面反射率（Ｒｅｌａｔｉｖｅ ｓｕｒｆａｃｅ ｒｅｆｌｅｃｔａｎｃｅ）、Ｓ_Ｋは準拠領域の各ピクセルのＫ色チャネルにおいて物体表面の固有色を決める表面反射率、および

はＫ色チャネルにおけるチャネル内比率尺度、ＫはＲまたはＧまたはＢ色チャネルである。

Here, the relative surface reflectance in the K color channel of each pixel of _{P K} compliant region (Relative surface reflectance), _{S K} is the surface reflectance that determines the intrinsic color of the object surface in the K color channel of each pixel of the criterion region ,and

Is the intra-channel ratio measure in the K color channel, and K is the R, G, or B color channel.

  At this time, the reason why the relative surface reflectance is calculated is that the surface reflectance that determines the intrinsic color of the object surface cannot be calculated. Therefore, the relative surface reflectance calculated by the mathematical formula 3 is used as the surface reflectance. This is to estimate the approximate reflectance.

In S <b> 40, the color restoration unit 50 restores the inherent surface color for each pixel in the compliant region by multiplying the luminance for each pixel in the compliant region by the relative surface reflectance for each color channel.
The luminous intensity of each pixel in the reference area is a sum of values indicated by the product of the surface reflectance and the illumination intensity for each color channel of each pixel in the reference area. This is done by mathematical formula 4.

[Equation 4]
L = V _R + V _G + V _B
(Mathematical formula 4)

Here, L is pixel-specific intensity, V _R is the hue value at the R color channel of each pixel constituting the compliance area, V _G is the hue value at the G color channel of each pixel constituting the compliance area, and V _B is It is a hue value in the B color channel of each pixel constituting the reference region.
In the compliant region, the restored surface color of each pixel can be obtained from the light intensity derived from Equation 4 and the surface reflectance derived from Equation 5 below. Equation 5 is a formula for calculating the intrinsic surface color for each pixel.

Here, the relative in K color channel of each pixel C _K unique surface color reconstructed in K color channel of each pixel constituting the compliance area, L is the pixel-specific intensity, and P _K constituting the criterion region The surface reflectance.
In S50, the adjustment coefficient calculation unit 60 calculates an adjustment coefficient for each color channel in the compliant area. The adjustment coefficient is a coefficient for eliminating the effect of the light source in the color restoration process of the input video, and the adjustment coefficient is calculated by the following mathematical formula 6.

ここで、ｍ_Ｋは準拠領域のＫ色チャネルの調整係数、

は準拠領域に属したピクセルが有するＫ色チャネルの固有表面色の和、および

は準拠領域に属したピクセルが有するＫ色チャネルにおける色相値Ｖ_Ｋの和である。

Where m _K is the adjustment factor for the K color channel in the compliant region,

Is the sum of the intrinsic surface colors of the K color channels of pixels belonging to the compliant region, and

Is the sum of the hue values V _K in the K color channel of the pixels belonging to the reference area.

In S60, the video restoration unit 70 restores the color of the input video using the adjustment coefficient calculated in S50.
The color restoration of the inputted video is performed by the following mathematical formula 7.

[Equation 7]
RC _K = V _K × m _K
(Formula 7)

Here, RC _K is the restored hue value in the K color channel of each pixel constituting the input video, V _K is the hue value in the K color channel of each pixel of the input video image, and m _K is It is an adjustment coefficient in the K color channel in the compliant area.

  In S70, the first correction unit 73 determines whether or not the restored hue value exceeds the limit value set to 0 to 255, and then the restored hue value does not exceed the limit value. If so, exit.

If the restored hue value exceeds the limit value, the first correction unit 73 performs overflow control in S73, and corrects the restored hue value to be equal to or less than the limit value.
The overflow control is performed by the following mathematical formula 8.

Here, RC ′ _K is a corrected hue value in the K color channel of each pixel constituting the restored video, max _original is a maximum hue value before restoration of the inputted video image, and max _recovered is inputted. This is the maximum hue value after restoration of the video image.

  Since the output value for each color channel of the video that can be expressed on the monitor is 256 levels from 0 to 255, when the restored hue value exceeds the limit value 255, the mathematical expression 8 is used and the color restoration system is used. By adjusting the output value of the hue value restored by 10, the restored hue value can be corrected below the limit value.

In S75, the second correction unit 75 recorrects the corrected hue value by multiplying the inherent surface color corrected in S73 by an output modulation factor (Gain factor).
The re-correction for the corrected hue value is performed by the following mathematical formula 9.

ここで、ＲＣ”_Ｋは各ピクセルのＫ色チャネルにおける再補正された色相値、Ｇ_ＫはＫ色チャネルにおける出力変調要因（Ｇａｉｎ ｆａｃｔｏｒ）、

は復元された映像イメージのＫ色チャネルにおけるピクセル別の色相値の和、

は補正された映像イメージのＫ色チャネルにおけるピクセル別の色相値の和、および

は閾値化値（Ｔｈｒｅｓｈｏｌｄｉｎｇ ｖａｌｕｅ）であって、ピクセルのＫ色チャネルにおける最大限界値である２５５である。

Here, RC ″ _K is a recorrected hue value in the K color channel of each pixel, G _K is an output modulation factor (Gain factor) in the K color channel,

Is the sum of the hue values for each pixel in the K color channel of the restored video image,

Is the sum of pixel-specific hue values in the K color channel of the corrected video image, and

Is a thresholding value, which is 255, the maximum limit value in the K color channel of the pixel.

  If the product of the corrected hue value and the output modulation factor does not exceed the threshold value as shown in the mathematical formula 9, the product of the corrected hue value and the output modulation factor is corrected again. If the product of the corrected hue value and the output modulation factor exceeds the threshold value, the threshold value can be a re-corrected hue value.

  FIG. 3 is a reference diagram simply showing the color restoration process of the color restoration system according to the present invention. As shown in FIG. 3, the color restoration process for the input image includes a process of dividing or redistributing the luminous intensity (for example, the total amount of light energy) of each pixel in consideration of the hue value for each of the three channels in which distortion can be ignored. Can be considered. At this time, the hue values for the three channels for which distortion can be ignored can be obtained by multiplying the adjustment coefficient obtained from the calculation process of the approximate value based on the in-channel ratio scale. The color restoration process of the color restoration system thus realized has an advantage that is very similar to the principle of realizing the color constancy of the human visual system.

  FIG. 4 is a reference diagram for a result of restoring an image damaged by red illumination using a color restoration system according to a preferred embodiment of the present invention. As shown in FIG. 4, an image (a) taken under standard illumination around noon was changed (b) to a state damaged by red illumination for the test of the color restoration system 10.

Further, the image (b) was generated as an image (c) restored using the color restoration system 10 and an image (d) restored using a conventional color restoration program (Adobe Photoshop).
As a result of comparison between the (c) image and the (d) image, the (c) image was restored to be substantially similar to the (a) image compared to the (d) image.

  FIG. 5 is a reference diagram for a result of restoring an image damaged by green illumination using a color restoration system according to an exemplary embodiment of the present invention. As shown in FIG. 5, an image (a) taken under standard illumination around noon was changed (b) to a state damaged by green illumination for the test of the color restoration system 10.

Further, the image (b) was generated as an image (c) restored using the color restoration system 10 and an image (d) restored using a conventional color restoration program (Adobe Photoshop).
As a result of comparison between the (c) image and the (d) image, the (c) image was restored to be substantially similar to the (a) image compared to the (d) image.

  FIG. 6 is a reference diagram for a result of restoring an image damaged by blue illumination using a color restoration system according to a preferred embodiment of the present invention. As shown in FIG. 6, an image (a) taken under standard illumination around noon was changed (b) to a state damaged by blue illumination for the test of the color restoration system 10.

Further, the image (b) was generated as an image (c) restored using the color restoration system 10 and an image (d) restored using a conventional color restoration program (Adobe Photoshop).
As a result of comparison between the (c) image and the (d) image, the (c) image was restored to be substantially similar to the (a) image compared to the (d) image.

  Although not shown, the present invention provides a computer-readable recording medium on which information coded by an instruction word that can be executed in a computer is recorded. Examples of the computer-readable recording medium include ROM, RAM, CD-ROM, magnetic tape, floppy disk, optical data storage device and the like. Programs, codes, and segments that can realize each function for realizing the recording medium can be easily inferred by a programmer in the technical field to which the present invention belongs.

  The above description is merely illustrative of the technical idea of the present invention, and any person who has ordinary knowledge in the technical field to which the present invention belongs can be used without departing from the essential characteristics of the present invention. Various modifications, changes and replacements are possible. Accordingly, the embodiments and the accompanying drawings disclosed in the present invention are not intended to limit the technical idea of the present invention, but are for explanation, and the technical idea of the present invention is described by such an embodiment and the accompanying drawings. The range is not limited. The protection scope of the present invention should be construed in accordance with the scope of claims, and all technical ideas within the equivalent scope should be construed as being included in the scope of rights of the present invention.

  The present invention can restore an image whose color is distorted by multiple light sources closer to the unique color, and overcomes the unique color problem that the unique color is not restored when the bias for the specific color exceeds the threshold. Therefore, it can be used for a digital camera, a video apparatus, and a video processing program in the form of a chip or a program.

DESCRIPTION OF SYMBOLS 10: Color restoration system 20: Reference area determination part 30: Ratio scale calculation part 40: Surface reflectance calculation part 50: Color restoration part 60: Adjustment coefficient calculation part 70: Image restoration part 73: 1st correction part 75: 2nd Correction unit

Claims (18)

a) determining a compliant area in consideration of a dispersion value for each color channel of the video to be restored;
b) calculating a relative surface reflectivity for the defined compliant region, and using the calculated relative surface reflectivity to calculate an adjustment factor for eliminating light source distortion effects; and c) the adjustment factor A color restoration method comprising the step of restoring the color of the video to be restored by using the method. Said step b)
b1) calculating an in-channel ratio measure (IR) for each pixel of the reference region;
b2) calculating a relative surface reflectance (P) of the compliant region using the intra-channel ratio measure (IR);
b3) restoring the intrinsic surface color (C) using the luminous intensity (L) of the reference area and the relative surface reflectance (P); and b4) the intrinsic surface color (C) of the reference area and the 2. The color restoration method according to claim 1, further comprising a step of calculating an adjustment coefficient for eliminating a light source distortion effect using a hue value for each channel in the reference region.   2. The color restoration method according to claim 1, wherein in the step a), the reference area is an area having a maximum sum of dispersion values for each color channel of the video to be restored. Said step a)
a1) dividing the video to be restored into a plurality of sub-videos;
a2) calculating a sum of dispersion values in each color channel for each of the divided sub-pictures, and setting a sub-picture having the maximum sum of dispersion values as a first sub-picture;
a3) subdividing the first sub-picture into a plurality of sub-pictures;
a4) calculating a sum of dispersion values in each color channel for each of the subdivided sub-pictures, and setting a sub-picture that maximizes the sum of the dispersion values as a second sub-picture;
a5) comparing the sum of variance values according to the first sub-picture with the sum of variance values according to the second sub-picture; and a6) depending on the result of the comparison, the first sub-picture into the compliant area. The method according to claim 1, further comprising the step of repeatedly executing the steps corresponding to the steps a3) to a5) for the second sub-picture. In the step a6), repeatedly executing the steps corresponding to the steps a3) to a5) means that the maximum value of the sum of the variance values for each video channel in the subdivided sub-video is the value before the subdivision. The above steps are executed until the sum of the variance values of the video channels for the sub-videos is smaller than or equal to the sum.
The color restoration method according to claim 4, wherein the reference area defines the sub-video before the subdivision as a reference area. Said step a)
a11) extracting sub-images by applying windows of different sizes to the image to be restored;
a22) calculating a sum of dispersion values for each color channel for each of the extracted sub-pictures; and a33) defining a sub-picture having a maximum sum of dispersion values for each color channel as a reference area. The color restoration method according to claim 1, wherein: In step b1)
3. The intra-channel ratio scale (IR) is a ratio of a hue value for each color channel of each pixel constituting the reference area to a hue value for each color channel of the reference area. Color restoration method. In the step b2), the relative surface reflectance is expressed by the following equation 3.

Where P _K is the relative surface reflectance in the K color channel of each pixel in the compliant region,

3. The color restoration method according to claim 2, wherein is an intra-channel ratio scale in the K color channel, and K is an R, G, or B color channel. In the step b3), the luminous intensity is obtained by the following equation 4.
[Equation 2]
L = V _R + V _G + V _B
(Formula 4)
Here, L is pixel-specific intensity, V _R is the hue value at the R color channel of each pixel constituting the compliance area, V _G is the hue value at the G color channel of each pixel constituting the compliance area, and V _B is The color restoration method according to claim 2, wherein the color value is a hue value in a B color channel of each pixel constituting the reference area. In the step b4), the adjustment coefficient is expressed by the following equation 6.

Here, m _K is the adjustment coefficient of the K color channel in the compliant region,

Is the sum of the intrinsic surface colors of the K color channels of the pixels belonging to the reference region,

The color restoration method according to claim 2, wherein is a sum of hue values V _K in K color channels of pixels belonging to the reference area. Following step c) above,
d1) when the hue value for each color channel of each pixel of the restored video exceeds a limit value, executing overflow control to correct the value to be equal to or less than the limit value; and d2) using an output modulation factor, The color restoration method according to claim 1, further comprising the step of recorrecting the hue value of each corrected color channel of each pixel.   12. The method of claim 11, wherein in the step d1), the overflow control uses a ratio value between a maximum hue value before restoration of the input video image and a maximum hue value after restoration. Color restoration method.   12. The color restoration method according to claim 11, wherein in the step d2), the re-corrected hue value is a product of the output modulation factor and the corrected specific surface color for each color channel. A computer-readable recording medium on which information coded by an instruction word executable in a computer is recorded for causing the computer to execute the color restoration method according to any one of claims 1 to 13. A compliant region determining unit that determines a region in which the sum of dispersion values of each color channel of the video to be restored is maximized as a compliant region;
An adjustment coefficient calculation unit that calculates an adjustment coefficient for eliminating a distortion effect of a light source in the compliant region; and a video restoration unit that restores the color of the video to be restored using the adjustment coefficient. Color restoration system. A ratio scale calculator for calculating an in-channel ratio scale (IR) of the compliant region;
A surface reflectance calculator that calculates a relative surface reflectance (P) of the reference region using the in-channel ratio scale (IR); and a luminous intensity (L) of the reference region and the relative surface reflectance (P) ) To restore the unique surface color (C) for each pixel,
The adjustment coefficient calculation unit calculates an adjustment coefficient for eliminating a distortion effect of a light source using the intrinsic surface color (C) of the compliant area and a hue value for each channel in the compliant area. The color restoration system according to claim 15.   A first correction unit configured to correct the hue value for each color channel to be equal to or less than the predetermined limit value through overflow control when the hue value for each color channel of the image restored by the image restoration unit exceeds a predetermined limit value; The color restoration system according to claim 15, further comprising:   The image processing apparatus further includes a second correction unit that re-corrects the corrected hue value for each color channel by multiplying the hue value for each color channel corrected by the first correction unit and an output modulation factor. Item 18. The color restoration system according to Item 17.

Images