CN111612702B - Neutral color correction post-treatment method for color migration - Google Patents

Neutral color correction post-treatment method for color migration Download PDF

Info

Publication number
CN111612702B
CN111612702B CN202010263116.XA CN202010263116A CN111612702B CN 111612702 B CN111612702 B CN 111612702B CN 202010263116 A CN202010263116 A CN 202010263116A CN 111612702 B CN111612702 B CN 111612702B
Authority
CN
China
Prior art keywords
image
color
saturation
correction
neutral
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.)
Active
Application number
CN202010263116.XA
Other languages
Chinese (zh)
Other versions
CN111612702A (en
Inventor
沈徐铭
邹安娥
张显斗
王勇
陈伟贲
李思远
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Hangzhou Dianzi University
Original Assignee
Hangzhou Dianzi University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Hangzhou Dianzi University filed Critical Hangzhou Dianzi University
Priority to CN202010263116.XA priority Critical patent/CN111612702B/en
Publication of CN111612702A publication Critical patent/CN111612702A/en
Application granted granted Critical
Publication of CN111612702B publication Critical patent/CN111612702B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T5/00—Image enhancement or restoration
    • G06T5/90—Dynamic range modification of images or parts thereof
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10004—Still image; Photographic image
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10024—Color image
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20212—Image combination
    • G06T2207/20224—Image subtraction

Landscapes

  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Facsimile Image Signal Circuits (AREA)
  • Color Image Communication Systems (AREA)
  • Image Processing (AREA)

Abstract

The invention discloses a neutral color correction post-treatment method for color migration. The method comprises the following steps: step (1) calculating L of the original image * a * b * A saturation map under color space; step (2) pairThe saturation diagram obtained in the step (1) is standardized and enhanced; step (3) calculating L of the original image and the migration image * a * b * Three-way difference values are obtained to obtain a difference image; step (4) according to the neutral color correction intention, carrying out saturation weighting on the difference image in the step (3), and calculating to obtain a neutral color correction image; step (5) carrying out tone correction on the neutral color correction image to obtain a final correction image; the final rectified image is then converted back to the sRGB color space for storage and display. According to the invention, through chromaticity saturation information calculation, on the premise of ensuring a good overall color migration effect, the neutral color which is originally color-shifted after color migration is corrected, so that the visual quality of an image after color migration is improved.

Description

Neutral color correction post-treatment method for color migration
Technical Field
The invention relates to a neutral color correction post-treatment method for color migration. Belongs to the technical fields of computer vision, image fusion, color correction and the like.
Background
Color migration is a popular problem in the field of computer vision, and by transferring the color style of a reference image onto an original image and by different reference images, the color style can be changed on the premise of not changing the content of the original image, so as to simulate different illumination, weather conditions, scene materials and even artistic color effects.
The research of color migration is focused on how to realize more accurate transmission and more similar color style conversion, such as color migration based on texture, color migration based on advanced semantic correspondence, color migration based on local area, artifact caused by color migration, and the like.
In the current color migration approach, however, the weighting of neutral colors (black, white, gray) or their retention is not considered. When the human eye vision evaluates the image quality, the sensitivity of the neutral color is higher than that of the region with obvious information of the non-neutral color, and if the neutral color is changed in chromaticity before and after migration, the visual quality of the image is obviously reduced, as shown in fig. 1.
Disclosure of Invention
The technical scheme adopted by the invention for solving the practical application problem is a neutral color correction post-treatment method aiming at various existing color migration methods, and the treatment process is that the treatment process is in L * a * b * The color space comprises the following specific steps:
step (1) calculating L of the original image * a * b * A saturation map under color space;
step (2) carrying out standardization and enhancement treatment on the saturation map obtained in the step (1);
step (3) calculating L of the original image and the migration image * a * b * Three-way difference values are obtained to obtain a difference image;
the original image and the migration image are images with different color styles and the same content;
step (4) according to the neutral color correction intention, carrying out saturation weighting on the difference image in the step (3) to calculate a neutral color correction image;
step (5) carrying out tone correction on the neutral color correction image to obtain a final correction image; the final rectified image is then converted back to the sRGB color space for storage and display.
Further, the default color space of the image stored in the step (1) is the sRGB color space, and sRGB color space is equal to L * a * b * Firstly, inverse gamma correction is needed, and the calculation method is as follows:
Figure BDA0002440167400000021
where rgb is the three channel value of the sRGB color space, respectively, with a dynamic range of 0, 255. The gamma (x) function is calculated as follows:
Figure BDA0002440167400000022
direct conversion to L by RGB * a * b * Space, the middle requires a transformation through XYZ space. The calculation method of converting XYZ from RGB is as follows:
Figure BDA0002440167400000023
l can be realized by XYZ space * a * b * The calculation method is as follows:
Figure BDA0002440167400000024
wherein ,Xn Y n Z n The standard white XYZ value is usually the standard white tristimulus value under the illumination condition of a D65 light source n Y n Z n ，(X n ，Y n ，Z n )＝（0.95047，1，1.08883)。
And the function f (t) is calculated as follows:
Figure BDA0002440167400000025
the saturation map S of the original image is calculated as follows:
Figure BDA0002440167400000031
further, in step (2), in order to enable the low-saturation image to still retain most of the color migration effect after the neutral color correction, the saturation map S needs to be normalized, so as to obtain a normalized saturation map S', which is as follows:
Figure BDA0002440167400000032
wherein ,Smin Is the minimum value of all pixels in the saturation diagram S, and S max Is the maximum value of all pixels in the saturation map S.
To further ensure the color migration effect, the saturation value of the saturation specific percentile is used for further weighting to realize enhancement, and an enhanced saturation diagram S is obtained * The method is characterized by comprising the following steps:
Figure BDA0002440167400000033
wherein ,Sn And (3) after the right small to large order in the normalized saturation diagram is shown, the normalized saturation value at the n% is taken as 30, namely the normalized saturation diagram is ordered from small to large to 30% of the saturation value.
Since the saturation is higher than 1 after the processing of formula (8), the outlier processing is required as follows:
Figure BDA0002440167400000034
for the final saturation map
Figure BDA0002440167400000035
The convolution kernel is again Gaussian filtered to size x size, which is determined by saturation map +.>
Figure BDA0002440167400000036
Is determined by the size of m and n in saturation diagram +.>
Figure BDA0002440167400000037
Wherein round () is a rounding function; the method comprises the following steps:
Figure BDA0002440167400000038
further, in step (3), L of the original image and the migration image is calculated * a * b * Three-way difference values are obtained to obtain a difference image, and the three-way difference image is specifically as follows:
Figure BDA0002440167400000039
wherein ,
Figure BDA0002440167400000041
l for migrating images * a * b * Three channel value->
Figure BDA0002440167400000042
L as original image * a * b * Three channel values. Note that Δl * Δa * Δb * Distinguishing between positive and negative values.
Further, in the step (4), a color shift image of the neutral color correction is calculated according to the neutral color correction intention.
If the neutral correction is intended to be fully preserved for neutral, the correction calculation method is as follows:
Figure BDA0002440167400000043
if the neutral correction is intended to be for neutral chromaticity retention (complete shift of luminance), the correction calculation method is as follows:
Figure BDA0002440167400000044
further, in step (5), in order to ensure color consistency before and after the correction of the neutral color, the color tone correction is required to be performed, a final corrected image is obtained, and the solving targets are as follows:
Figure BDA0002440167400000045
wherein ,
Figure BDA0002440167400000046
the solution is as follows:
Figure BDA0002440167400000047
wherein
Figure BDA0002440167400000048
wherein
Figure BDA0002440167400000049
Finally, to
Figure BDA0002440167400000051
Lab space three-channel values as final corrected image, the final corrected image is converted from L by inversely implementing the color space in step (1) * a * b * And converting the color space back to the sRGB color space to obtain a color migration image corrected by neutral color.
The technical scheme provided by the invention has the beneficial effects that:
aiming at the problems of serious color cast, influence on image quality, visual experience and the like of the traditional color migration method, the invention corrects the seriously-cast neutral color in the image after color migration on the premise of realizing color style conversion on the premise of ensuring better overall color migration effect by calculating the chroma saturation information, thereby improving the visual quality of the image after color migration.
Drawings
FIG. 1 is a diagram showing a neutral color shift problem in the result of the conventional image migration method;
FIG. 2 is an original image in an embodiment;
FIG. 3 is a migration image of neutral color cast obtained by the Reinhard method in the example;
FIG. 4 is a saturation diagram of an original image in an embodiment;
FIG. 5 is a normalized saturation map obtained by normalizing the original image saturation in the embodiment;
FIG. 6 is a result of overall enhancement of the saturation map using a normalized saturation level at 30% from the small to large percentile in the example;
FIG. 7 is a final saturation map obtained by smoothing the saturation map of FIG. 6 through a Gaussian filter kernel convolution in an embodiment;
FIG. 8 is a color shift image with neutral chromaticity preserved corrected;
fig. 9 is a color shift image with neutral color fully preserved and corrected.
Detailed Description
The invention is further described below with reference to the drawings and examples.
The invention mainly provides a neutral color correction post-treatment method for color migration. And judging the probability of neutral color of each pixel of the original image according to the saturation information of each pixel of the original image for the migrated image and the original image after the color migration, and calculating the algorithm super-parameters according to the saturation information through a continuous and smooth prior function, wherein the super-parameters determine the degree of the color migration. For the pixels with medium and high saturation, performing high-degree color migration; for low saturation pixels, a lower degree of migration is performed. Whereas the neutral color region tends to be of low saturation, whereby a neutral color correction post-treatment for color migration can be achieved.
Embodiments of the present invention are common landscape images. Referring to fig. 2 (original image) and fig. 3 (color migration image obtained by Reinhard method), the process of the embodiment of the present invention includes the following steps:
step (1) calculating L of the original image * a * b * A saturation map under color space;
step (2) carrying out standardization and enhancement treatment on the saturation map obtained in the step (1);
step (3) calculating L of the original image and the migration image * a * b * Three-way difference values are obtained to obtain a difference image;
step (4) according to the neutral color correction intention, carrying out saturation weighting on the difference image in the step (3), and calculating to obtain a neutral color correction image;
step (5) carrying out tone correction on the neutral color correction image to obtain a final correction image; the final rectified image is then converted back to the sRGB color space for storage and display.
In step (1), taking fig. 2 as an example of an original image, the image is calculated to be L * a * b * The saturation image in color space is shown in fig. 4. Since the default stored color space of the image is the sRGB color space, sRGB color space through L * a * b * Firstly, inverse gamma correction is needed, and the calculation method is as follows:
Figure BDA0002440167400000061
where rgb is the three channel value of the sRGB color space, respectively, with a dynamic range of 0, 255. The gamma (x) function is calculated as follows:
Figure BDA0002440167400000062
direct conversion to L by RGB * a * b * Space, the middle requires a transformation through XYZ space. The calculation method of converting XYZ from RGB is as follows:
Figure BDA0002440167400000063
by XYZ we can realize the transformation of L * a * b * The calculation method is as follows:
Figure BDA0002440167400000071
wherein ,Xn Y n Z n The standard white XYZ value is usually the standard white tristimulus value under the illumination condition of a D65 light source n Y n Z n ，(X n ，Y n ，Z n )＝(0.95047，1，1.08883)。
And the function f (t) is calculated as follows:
Figure BDA0002440167400000072
the original image saturation information is calculated as follows:
Figure BDA0002440167400000073
in step (2), in order to enable the low-saturation image to still retain most of the color migration effect after the neutral color correction, the saturation map S needs to be normalized, so as to obtain a normalized saturation map S', as shown in fig. 5, the process is as follows:
Figure BDA0002440167400000074
wherein ,Smin Is the minimum value of all pixels in the saturation diagram S, and S max Is the maximum value of all pixels in the saturation diagram S;
in order to further ensure the color migration effect, the saturation value of the saturation specific percentile is used for further weighting to realize enhancement, and an enhanced saturation diagram S is obtained * As shown in fig. 6, the specific steps are as follows:
Figure BDA0002440167400000075
wherein ,Sn Representing normalized post-saturationAfter the right small to large order in the sum degree diagram, the normalized saturation value at the n% is taken as 30, namely the normalized saturation diagram is ordered from small to large to 30% saturation value;
since the saturation is higher than 1 after the processing of formula (8), the outlier processing is required as follows:
Figure BDA0002440167400000081
for the final saturation map
Figure BDA0002440167400000082
Performing Gaussian filtering with convolution kernel size of size multiplied by size again to obtain a filtered image as shown in fig. 7; size is defined by saturation map->
Figure BDA0002440167400000083
Is determined by the size of m and n in saturation diagram +.>
Figure BDA0002440167400000084
Wherein round () is a rounding function; the method comprises the following steps:
Figure BDA0002440167400000085
calculating L of original image and migration image in step (3) * a * b * Three-way difference values are obtained to obtain a difference image, and the three-way difference image is specifically as follows:
Figure BDA0002440167400000086
wherein ,
Figure BDA0002440167400000087
l for migrating images * a * b * Three channel value->
Figure BDA0002440167400000088
L as original image * a * b * Three channel values. Note that Δl * Δa * Δb * Distinguishing between positive and negative values.
And (4) calculating a color migration image of the neutral color correction according to the neutral color correction intention.
If the neutral color correction is intended to be completely preserved for the neutral color, the obtained neutral color correction image is as shown in fig. 9, the correction calculation method is as follows:
Figure BDA0002440167400000089
if the neutral correction is intended to be a color retention for neutral, i.e., the brightness is completely shifted, the resulting neutral corrected image is as shown in fig. 8, and the correction calculation method is as follows:
Figure BDA0002440167400000091
in the step (5), in order to ensure the consistency of the color before the correction of the neutral color, the tone correction is required to be carried out, so that a final corrected image is obtained; converting the tone correction into a form of target solution, and specifically solving the target as follows:
Figure BDA0002440167400000092
wherein ,
Figure BDA0002440167400000093
the solution is as follows:
Figure BDA0002440167400000094
wherein
Figure BDA0002440167400000095
wherein
Figure BDA0002440167400000096
Finally, to
Figure BDA0002440167400000097
Lab space three-channel values as final corrected image, the final corrected image is converted from L by inversely implementing the color space in step (1) * a * b * And converting the color space back to the sRGB color space to obtain a color migration image corrected by neutral color.
Thereby, L of the neutral color corrected image is obtained * a * b * The characterization under space also requires an inverse transformation back to the sRGB color space.
Finally, the final rectified image is transformed from L by inversely implementing the color space conversion in step (1) * a * b * The color space is converted back to sRGB color space, thus obtaining a neutral color corrected Reinhard color migration image, and L is calculated * a * b * The first step of the conversion of the color space back to XYZ color space is as follows:
Figure BDA0002440167400000101
wherein the inverse solution of f (x) is as follows:
Figure BDA0002440167400000102
the calculation of the conversion back to XYZ color space is as follows:
Figure BDA0002440167400000103
wherein ,(Xn ，Y n ，Z n )＝(0.95047，1，1.08883)。
Whereas the first conversion of XYZ color space to sRGB color space is as follows:
Figure BDA0002440167400000104
through the inverse gamma correction, the conversion to the sRGB color space can be completed.
Figure BDA0002440167400000105
Wherein, the inverse gamma correction function is as follows:
Figure BDA0002440167400000106
thereby, color correction of the color shift image based on the original image saturation information is completed.
The foregoing is a further detailed description of the invention in connection with the preferred embodiments, and is not intended to limit the practice of the invention to such description. It will be understood by those skilled in the art that various changes in detail may be effected therein without departing from the scope of the invention as defined by the claims appended hereto.

Claims (4)

1. A neutral color correction post-treatment method for color migration is characterized by comprising the following steps:
step (1) calculating L of the original image * a * b * A saturation map under color space;
step (2) carrying out standardization and enhancement treatment on the saturation map obtained in the step (1);
step (3) calculating L of the original image and the migration image * a * b * Three-channel difference values to obtain a difference image；
Step (4) according to the neutral color correction intention, carrying out saturation weighting on the difference image in the step (3), and calculating to obtain a neutral color correction image;
step (5) carrying out tone correction on the neutral color correction image to obtain a final correction image; then converting the final corrected image back to the sRGB color space for storage and display;
in step (2), in order to enable the low-saturation image to still retain most of the color migration effect after the neutral color correction, the saturation map S needs to be normalized, so as to obtain a normalized saturation map S', which is as follows:
Figure FDA0004090936770000011
wherein ,Smin Is the minimum value of all pixels in the saturation diagram S, and S max Is the maximum value of all pixels in the saturation diagram S;
to further ensure the color migration effect, the saturation value of the saturation specific percentile is used for further weighting to realize enhancement, and an enhanced saturation diagram S is obtained * The method is characterized by comprising the following steps:
Figure FDA0004090936770000012
wherein ,Sn The normalized saturation value at the n% after the normalized saturation diagram is ordered from small to large is represented, and n is 30, namely the normalized saturation diagram is ordered from small to large to 30%;
since the saturation is higher than 1 after the saturation enhancement formula processing, outlier processing is required, as follows:
Figure FDA0004090936770000013
for the final saturation map
Figure FDA0004090936770000014
The convolution kernel is again Gaussian filtered to size x size, which is determined by saturation map +.>
Figure FDA0004090936770000015
Is determined by the size of m and n in saturation diagram +.>
Figure FDA0004090936770000016
Wherein round () is a rounding function; the method comprises the following steps:
Figure FDA0004090936770000021
in the step (4), according to the intention of the neutral color correction, calculating a color migration image of the neutral color correction, specifically as follows:
if the neutral correction is intended to be fully preserved for neutral, the correction calculation method is as follows:
Figure FDA0004090936770000022
if the neutral correction is intended to be chromaticity preserving for neutral, i.e. brightness is completely shifted, the correction calculation method is as follows:
Figure FDA0004090936770000023
2. the method of post-correction processing for color migration of neutral colors according to claim 1, wherein L of the original image and the migrated image is calculated in the step (3) * a * b * Three-way difference values are obtained to obtain a difference image, and the three-way difference image is specifically as follows:
Figure FDA0004090936770000024
wherein ,
Figure FDA0004090936770000025
l for migrating images * a * b * Three channel value->
Figure FDA0004090936770000028
L as original image * a * b * Three channel values; note that Δl * Δa * Δb * Distinguishing between positive and negative values.
3. The method according to claim 2, wherein in step (5), in order to ensure consistency of colors before the neutral color correction, a tone correction is required to obtain a final corrected image; converting the tone correction into a form of target solution, and specifically solving the target as follows:
Figure FDA0004090936770000026
wherein ,
Figure FDA0004090936770000027
the solution is as follows:
Figure FDA0004090936770000031
wherein
Figure FDA0004090936770000032
wherein
Figure FDA0004090936770000033
Finally, to
Figure FDA0004090936770000034
L as final corrected image * a * b * Spatial three-channel values, the final rectified image is transformed from L by inversely implementing the color space conversion in step (1) * a * b * And converting the color space back to the sRGB color space to obtain a color migration image corrected by neutral color.
4. A method of post-treatment for color shifting neutral color correction according to claim 1 or 3, wherein step (1) is specifically implemented as follows:
since the default stored color space of the image is the sRGB color space, sRGB color space through L * a * b * Firstly, inverse gamma correction is needed, and the calculation method is as follows:
Figure FDA0004090936770000035
wherein, rgb is the three channel value of sRGB color space, its dynamic range is [0,255]; the gamma (x) function is calculated as follows:
Figure FDA0004090936770000036
direct conversion to L by RGB * a * b * Space, the middle needs to be transformed by XYZ space; the calculation method of converting XYZ from RGB is as follows:
Figure FDA0004090936770000037
realizes the L by XYZ * a * b * The calculation method is as follows:
Figure FDA0004090936770000041
wherein ,Xn Y n Z n The standard white XYZ value is usually the standard white tristimulus value under the illumination condition of a D65 light source n Y n Z n ，(X n ,Y n ,Z n ) = (0.95047,1,1.08883); and the function f (t) is calculated as follows:
Figure FDA0004090936770000042
the saturation map of the original image is calculated as follows:
Figure FDA0004090936770000043
/>
CN202010263116.XA 2020-04-07 2020-04-07 Neutral color correction post-treatment method for color migration Active CN111612702B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010263116.XA CN111612702B (en) 2020-04-07 2020-04-07 Neutral color correction post-treatment method for color migration

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010263116.XA CN111612702B (en) 2020-04-07 2020-04-07 Neutral color correction post-treatment method for color migration

Publications (2)

Publication Number Publication Date
CN111612702A CN111612702A (en) 2020-09-01
CN111612702B true CN111612702B (en) 2023-04-21

Family

ID=72203667

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010263116.XA Active CN111612702B (en) 2020-04-07 2020-04-07 Neutral color correction post-treatment method for color migration

Country Status (1)

Country Link
CN (1) CN111612702B (en)

Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110728722A (en) * 2019-09-18 2020-01-24 苏宁云计算有限公司 Image color migration method and device, computer equipment and storage medium

Family Cites Families (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR20050015107A (en) * 2003-08-02 2005-02-21 삼성전자주식회사 Image enhancement apparatus and method thereof
KR100621414B1 (en) * 2004-06-09 2006-09-08 삼성전자주식회사 Saturation-adaptive image enhancement apparatus and method the same
JP4923987B2 (en) * 2006-11-30 2012-04-25 株式会社ニコン Image processing apparatus and image processing program for correcting image color
JP2010199659A (en) * 2009-02-23 2010-09-09 Panasonic Corp Image processing apparatus, and image processing method
CN101621702B (en) * 2009-07-30 2011-08-24 北京海尔集成电路设计有限公司 Method and device for automatically adjusting chroma and saturation
CN108416741B (en) * 2018-01-23 2021-10-12 浙江工商大学 Rapid image defogging method based on brightness contrast enhancement and saturation compensation
CN110276727B (en) * 2019-05-21 2020-06-05 杭州电子科技大学 Color vision disorder-oriented color image color enhancement method
CN110349097B (en) * 2019-06-17 2021-06-25 苏州华星光电技术有限公司 Color enhancement method for image significance and image processing device

Patent Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110728722A (en) * 2019-09-18 2020-01-24 苏宁云计算有限公司 Image color migration method and device, computer equipment and storage medium

Also Published As

Publication number Publication date
CN111612702A (en) 2020-09-01

Similar Documents

Publication Publication Date Title
CN106780368B (en) Underwater image enhancement method based on foreground model
CN103593830B (en) A kind of low illumination level video image enhancement
CN105915909B (en) A kind of high dynamic range images layered compression method
CN108711140B (en) Image brightness uniformity real-time recovery method based on inter-class variance description
CN106504212A (en) A kind of improved HSI spatial informations low-luminance color algorithm for image enhancement
CN106897981A (en) A kind of enhancement method of low-illumination image based on guiding filtering
CN110706172B (en) Low-illumination color image enhancement method based on adaptive chaotic particle swarm optimization
CN101930596A (en) Color constancy method in two steps under a kind of complex illumination
CN107820069B (en) Video monitoring equipment ISP debugging method
Jang et al. Adaptive color enhancement based on multi-scaled Retinex using local contrast of the input image
CN109325918B (en) Image processing method and device and computer storage medium
CN111968065A (en) Self-adaptive enhancement method for image with uneven brightness
CN111105359A (en) Tone mapping method for high dynamic range image
CN107358578B (en) Yin-yang face treatment method and device
CN113658043A (en) Image processing method, image processing device, electronic equipment and readable storage medium
US8565523B2 (en) Image content-based color balancing
CN110766622A (en) Underwater image enhancement method based on brightness discrimination and Gamma smoothing
WO2022120799A1 (en) Image processing method and apparatus, electronic device, and storage medium
WO2023241339A1 (en) Color cast correction method and apparatus, device, storage medium and program product
CN111028181B (en) Image enhancement processing method, device, equipment and storage medium
CN111612702B (en) Neutral color correction post-treatment method for color migration
CN109102473B (en) Method for improving color digital image quality
WO2012153661A1 (en) Image correction device, image correction display device, image correction method, program, and recording medium
Yu et al. Color constancy-based visibility enhancement in low-light conditions
CN112581390B (en) Image color enhancement method, device, equipment and readable storage medium

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant