CN105761292B - One kind is based on color transfer and modified image rendering methods - Google Patents
One kind is based on color transfer and modified image rendering methods Download PDFInfo
- Publication number
- CN105761292B CN105761292B CN201610112647.2A CN201610112647A CN105761292B CN 105761292 B CN105761292 B CN 105761292B CN 201610112647 A CN201610112647 A CN 201610112647A CN 105761292 B CN105761292 B CN 105761292B
- Authority
- CN
- China
- Prior art keywords
- pixel
- image
- target gray
- feature
- color
- 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
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T11/00—2D [Two Dimensional] image generation
- G06T11/001—Texturing; Colouring; Generation of texture or colour
Abstract
The invention discloses one kind based on color transfer and modified image rendering methods, including step：Space is converted, and obtains the component information of the point brightness of target gray image any pixel and coloration；Read any one pixel t in target gray image；Using non local method, the selected pixels point s in the coloured image of source, pixel s determine that the Variance feature of window and discrete Fourier transform feature window Variance feature determining with pixel t institutes and discrete Fourier transform feature are close；The chromatic value of pixel t is tentatively coloured according to the chromatic value of pixel s；Coupling total variation model is established, solving model obtains the optimal solution of chromatic value, will acquire optimal solution as revised chromatic value, brightness value remains unchanged；It repeats the above steps, each pixel t of target gray image is modified and exports revised image.Color space is inconsistent after the present invention can preferably overcome the problems, such as coloring, realizes automatically and rapidly accurate coloring.
Description
Technical field
The present invention relates to one kind based on color transfer and modified image rendering methods, belong to the technology neck of image processing
Domain.
Background technology
In information is transmitted and is exchanged, image is important communications media.In human visual system, color is to embody to regard
One of an important factor for feeling cognition.Since human eye can not show a candle to colour information to the susceptibility of half-tone information, therefore, it is desirable to gray scale
Image becomes coloured image to enhance visual effect, i.e. image colorant (image colorization) technology, also referred to as grey
Spend image colorization.Image colorant is the computer assisted procedures that color is added to still image or video sequence.Exist at present
The fields such as video display, medical treatment, space probation and industry, scientific research have a wide range of applications, while the research of image rendering methods
Always active, the challenging research topic of image procossing and one, applied mathematics field.
The image rendering methods of view-based access control model cognition in recent years can be divided mainly into two classes：One kind is expanded based on local color
The color method opened；Another kind is the color method shifted based on color.At present, in the color method research shifted based on color
Aspect, due in target gray image, the coloring treatment at each pixel is independent, therefore this traditional coloring side
Method usually will appear the problem of color is inconsistent, and is still far from perfect using the color method that partial differential equation model.How to allow
Color keeps Space Consistency (spatial consistency) in transfer process, is always that color transfer color method is ground
The a great problem studied carefully.Moreover, there has been no unified methods for the image rendering methods based on color transfer.
Invention content
The technical problems to be solved by the invention be to overcome the deficiencies of the prior art and provide it is a kind of based on color transfer and
Modified image rendering methods, solve the prior art during the image colorant at each pixel coloring treatment be it is independent,
And the problem of color is caused to cannot keep Space Consistency in transfer process, to realize the accurately and quickly coloring of image.
It is of the invention that above-mentioned technical problem is specifically solved using following technical scheme：
One kind is based on source coloured image and target gray figure based on color transfer and modified image rendering methods, this method
Picture, including step：
Step 1, space conversion：YCbCr space is selected, target gray is obtained by the conversion of rgb space to YCbCr space
The component information of any pixel point brightness and coloration in image, i.e. (Y, Cb, Cr)；
Step 2 reads any one pixel t (Y in target gray imaget,Cbt,Crt), centered on pixel t,
The window of L × L is opened, the Variance feature of window and discrete Fourier transform feature where determining the pixel；Utilize non local side
Method, the selected pixels point s (Y in the coloured image of sources,Cbs,Crs), it is desirable that by the pixel s windows determined Variance feature and
Discrete Fourier transform feature and the pixel t determining window Variance features of institute and discrete Fourier transform in target gray image are special
It levies close；According to the chromatic value (Cb of pixel s in the source coloured images,Crs) to the coloration of pixel t in target gray image
It is worth (Cbt,Crt) tentatively coloured, the step is repeated until each pixel of preliminary coloring target gray image；
Step 3, basis tentatively colour the chromatic value of pixel t and the original brightness of point in obtained target gray image
Value Yt, establish coupling total variation model；The optimal solution that coupling total variation model obtains pixel t chromatic values is solved, by what is acquired
The optimal solution of chromatic value is remained unchanged as revised chromatic value, brightness value；
Step 4, repeating said steps 3, are modified each pixel t of target gray image and will be revised
The image output of pixel composition, that is, complete the coloring to target gray image.
Further, as a preferred technical solution of the present invention, the step 1 is further included to selected YCbCr
Space is normalized.
Further, as a preferred technical solution of the present invention, the step 2 selected pixels point s is specifically included：
The sampling set being made of several pixels is chosen in the coloured image of source；
Calculate in the sampling set pixel the Variance feature of determining window and discrete Fourier transform feature, and select
Pixel similar in window Variance feature wherein determining with pixel t institutes and discrete Fourier transform feature is taken as feasible solution；
Any one solution is selected from several feasible solutions as target solution, and extracts the pixel color corresponding to the target solution
Angle value.
Further, as a preferred technical solution of the present invention, the step 3 solves coupling entirely using ADM algorithms
Variation model.
Further, as a preferred technical solution of the present invention, the ADM algorithms specifically include：It will the full change of coupling
Differential mode type is changed into the convex programming problem of the linear convex set contrained of standard；Alternating iteration, which is carried out, using ADM algorithms solves acquisition most
Excellent solution.
The present invention can generate following technique effect using above-mentioned technical proposal：
(1), it is provided by the present invention based on color transfer and modified image rendering methods, mainly in preliminary tinting stage
It is respectively established with color correct, in preliminary tinting stage, it is colored with source that target gray image is found out by non local method
Pixel similar in characteristics of image, and the chromatic value at the pixel is passed into gray image, so as to fulfill to gray level image
Preliminary coloring.In the color correct stage, the total variation model of a coupling is established, carries out regional area color diffusion, is corrected
Color in preliminary coloring process at the inconsistent pixel of color, to obtain the rgb space coordinate of each pixel and complete
Image output after color.To image colorant end of output since selecting image space, in the case where ensureing that color is not crossed the border,
It realizes the diffusion of regional area color, reaches the quick and precisely amendment purpose at the inconsistent point of color.The present invention can preferably overcome
The problem of color space is inconsistent after coloring, realizes automatic, the quick accurate coloring to target gray image, and acquired results are better than
The image rendering methods based on color transfer existing at present.
(2) and the method for the present invention is during model solution, first turns model using the Lagrangian method of augmentation
The convex programming problem of the linear convex set contrained of standard is turned to, is then solved with alternating direction multiplier ADM algorithms.Compared to tradition
Color algorithm, the convergence and coloring precision are obtained for and effectively improve.Meanwhile the algorithm has stronger robust
Property.
Description of the drawings
Fig. 1 is the schematic diagram based on color transfer and modified image rendering methods of the present invention.
Specific embodiment
Embodiments of the present invention are described with reference to the accompanying drawings of the specification.
It is based on as shown in Figure 1, the present invention devises a kind of shifted based on color with modified image rendering methods, this method
Source coloured image and target gray image, this method mainly include four parts：1. it is converted between color；2. preliminary coloring；It is 3. right
The inconsistent point in space carries out color correct；4. coloured image exports.To image colorant end of output since selecting image space,
In the case where ensureing that color is not crossed the border, realize the diffusion of regional area color, reach quick and precisely repairing at the inconsistent point of color
Positive purpose, each step is respectively as following：
Step 1, color space conversion；In color space selected section, suitable color space is selected --- YCbCr is empty
Between, by the conversion of rgb space to YCbCr space, obtain the component information of the point brightness of image any pixel and coloration, i.e., (Y,
Cb,Cr)；The present invention selects the YCbCr space that can preferably characterize brightness of image, chromaticity, improves coloration efficiency.For carrying
The color space matrix of taking-up, needs to be normalized, and Cb (or Cr) value of each chrominance channel is held at 16/255
Between~240/255.
Step 2, the requirement by designed image rendering methods, secondly the color of preliminary coloured part moves process,
Its principle is to read any one pixel t (Y in target gray imaget,Cbt,Crt), centered on pixel t, open L × L
Window, determine the Variance feature of window where the pixel and discrete Fourier transform feature；Using non local method, in source
Selected pixels point s (Y in coloured images,Cbs,Crs), it is desirable that Variance feature and discrete Fu by the pixel s windows determined
In the determining window Variance features of pixel t institutes and discrete Fourier transform feature phase in leaf transformation feature, with target gray image
Closely, the close degree can according to the corresponding numberical range of coloring required precision setting, if error range is [- 0.05,0.05], but
The present invention is not limited to the range, other numberical ranges may be equally applied in this method.Thus according to picture in the source coloured image
Chromatic value (the Cb of vegetarian refreshments ss,Crs) to the chromatic value (Cb of pixel t in target gray imaget,Crt) tentatively coloured, weight
The multiple step is until each pixel t of preliminary coloring target gray image.
The process of above-mentioned selected pixels point s is：It chooses first in the coloured image of source and is adopted by what several pixels formed
Sample set；Then, calculate it is described sampling set in pixel the Variance feature of determining window and discrete Fourier transform feature,
And pixel similar in selection window Variance features wherein determining with pixel t institutes and discrete Fourier transform feature is as feasible
Solution；Any one solution is selected from several feasible solutions as target solution, and extracts the pixel chromatic value corresponding to the target solution.
The specific implementation process of preliminary tinting steps is as follows in the method for the present invention：Source coloured image S is in YCbCr color spaces
Under coordinate be (YS,CbS,CrS), coordinates of the target gray image T under YCbCr color spaces is (YT,CbT,CrT), wherein
CbT,CrT=0.For any one pixel t in target gray image T, using non local method, find out so that YTWith YS
Feature it is closest when, the pixel s in corresponding source coloured image S, and by the chromatic value V of pixel s0=(Cb0,Cr0)
Chromatic value as pixel t, you can realize preliminary coloring.Specifically, to improve coloration efficiency, source coloured image is closed
The mesh generation of suitable size, using n point on grid as sub-sampling set D (Sn), the pixel point set of target gray image
For D (T).For pixel t ∈ D (T) and s ∈ D (Sn), their coordinates in the picture are t (x, y) and s (x', y') respectively,
With t, region (i.e. window) brightness value of L × L sizes centered on s is respectively：
PL(t)={ YT(x+k,y+l)},PL(s)={ YS(x'+k,y'+l)}
(1.1)
For all pixels point in the two regions, two main characteristics of image are chosen：Variance feature, can be preferably
Picture engraving texture features；Discrete Fourier transform feature (DFT), can preferably picture engraving architectural characteristic.Color transfer
Model is as follows：
f1(t,L),f1(s, L) is the Variance feature of pixel t and pixel s respectively：
f1(t, L)=Var (PL(t)),f1(s, L)=Var (PL(s)) (1.2)
f2(t, L, ξ), f2(s, L, ξ) is DFT (discrete Fourier transform) feature of pixel t and pixel s respectively：
f2(t, L, ξ)=| DFT (PL(t),ξ)|,f2(s, L, ξ)=| DFT (PL(s),ξ)| (1.3)
Wherein ξ=(ξ1,ξ2) represent frequency.
For pixel t ∈ D (T) and s ∈ D (Sn), the index of correlation under two kinds of features is defined respectively：
d1(t, s, L)=| f1(t,L)-f1(s,L)| (1.4)
F in formula1Represent the variance of certain area pixel set, f2Represent the DFT values of certain area pixel set.d1And d2
The variance in the region centered on pixel t and s and the difference of DFT values are represented respectively.For arbitrary pixel t, need to acquire
d1And d2Corresponding s coordinate values when minimum.For solving result is made more accurately to solve, the smaller variable h of numerical value is introduced>0, and turn
Index number problem is turned to, the coordinate value of s when solving the index maximum.
By solving above-mentioned two formula, it is able to obtain under two kinds of features, the corresponding chromatic values of pixel t：
And then obtain the corresponding YCbCr space coordinate values of pixel t, using two kinds of features respectively to target gray image into
The preliminary coloring of row：
WithSolving result may have multiple, the design proposes arbitrary selection respectively, and one of those is feasible
Solution is as the target solution based on Variance feature and DFT features.
Step 3, point mainly inconsistent to space carry out color correct, i.e., tentatively colour in obtained target gray image
The chromatic value of pixel t and the original brightness value Y of the pointt, establish coupling total variation model；Part is corrected using color diffusion to think
Think, the coupling total variation model of proposition is as follows：
The Ω is bounded image area, and a, b are constant, and a=16/255, b=240/255.
In this model, preceding paragraph is couples total variation regularization term, the consequent fidelity item initially to colour.Wherein：α1,α2＞ 0
For weight parameter, g is monotonic decreasing function, is set asT ∈ R, k are set as 5000.D is gradient operator, and Δ is
Laplace operator.It is defined as below respectively：
GσFor Gaussian kernel, this model uses σ=1.Y0Brightness value for target gray image.|ΔGσ*Y0| it can be effective
Portray the structural information in target image.|ΔGσ*Y0| color boundary is represented when larger, corresponding g values are smaller at this time, the coupling
Total variation item can effectively prevent color and occur more zone phenomenon in diffusion process.In fidelity item,Tentatively to colour
In stage, using two kinds of obtained chromatic values of different characteristic.In the case that the model can not cross the border ensureing color, realization office
Portion's field color diffusion, so as to reach the automatic amendment to color after tentatively colouring.
The optimal solution V of pixel t chromatic values is obtained by solving coupling total variation model, once optimal solution V is determined, with reference to
Brightness value Y0, you can the color at each pixel t is obtained, by the chromatic value of the optimal solution to pixel t after tentatively colouring
Chromatic value be modified, brightness value remains unchanged, and then can complete the automatic coloring of target image.It does not cross the border in guarantee color
In the case of, it realizes the diffusion of regional area color, reaches the quick and precisely modified purpose of the inconsistent point in space.
For above-mentioned model, present invention preferably uses following derivation algorithms：
1. pass through created symbol Γ ≡ { V=(Cb, Cr) | a≤Cb, Cr≤b } and auxiliary variableAnd W,
Model is changed into the convex programming problem of the linear convex set contrained of following standard：
2. using change of direction Multiplier Algorithm, i.e. ADM algorithms, Lagrangian is introducedObtain as
Lower Augmented Lagrangian Functions：
Algorithm flow is as follows：
Assuming thatIt is arbitrary real number with λ, μ ＞ 0, then ordered series of numbersIt is converged to by the calculating of ADM algorithmsWhereinIt is the optimal solution of object function (2.1).
The present invention, using ADMM algorithms, can effectively overcome traditional gradient descent method speed slow and hold during model solution
The shortcomings that being easily absorbed in local minimum.Compared to traditional stains algorithm, the convergence and coloring precision have been obtained for
Effect improves.Meanwhile the algorithm has higher robustness, but the present invention is not limited to this kind of derivation algorithms.
Step 4 repeats step 3, each pixel t of target gray image is modified, by revised pixel
The image output of point composition.Last image output portion using YCbCr to rgb color space inverse conversion, obtains target grey
In image the rgb space coordinate of each pixel t and complete coloring after image output.
To sum up, it is proposed by the invention based on color transfer and modified image rendering methods, it is opened from image space selection
Begin to image colorant end of output, in the case where ensureing that color is not crossed the border, realize the diffusion of regional area color, reach color not
Quick and precisely amendment purpose at consistent point.Color space is inconsistent after preferably overcoming the problems, such as coloring, realizes to target
Automatic, the quick accurate coloring of gray level image, acquired results are better than the image rendering methods based on color transfer existing at present.
Embodiments of the present invention are explained in detail above in conjunction with attached drawing, but the present invention is not limited to above-mentioned implementations
Mode, within the knowledge of a person skilled in the art, can also be under the premise of present inventive concept not be departed from
It makes a variety of changes.
Claims (5)
1. one kind is based on source coloured image and target gray figure based on color transfer and modified image rendering methods, this method
Picture, which is characterized in that including step：
Step 1, space conversion：YCbCr space is selected, target gray image is obtained by the conversion of rgb space to YCbCr space
Interior any pixel point brightness and the component information of coloration, i.e., (Y, Cb, Cr)；
Step 2 reads any one pixel t (Y in target gray imaget,Cbt,Crt), centered on pixel t, open L
The window of × L, the Variance feature of window and discrete Fourier transform feature where determining the pixel；Using non local method,
The selected pixels point s (Y in the coloured image of sources,Cbs,Crs), it is desirable that by the pixel s windows determined Variance feature and from
Dissipate Fourier transformation feature and the pixel t determining window Variance features of institute and discrete Fourier transform feature in target gray image
Under the index value asked for it is close；According to the chromatic value (Cb of pixel s in the source coloured images,Crs) in target gray image
Chromatic value (the Cb of pixel tt,Crt) tentatively coloured, the step is repeated until preliminary colour each of target gray image
Pixel；
Step 3, basis tentatively colour the chromatic value of pixel t and the original brightness value Y of the point in obtained target gray imaget,
Coupling total variation model is established, specially：
In formula, the Ω is bounded image area；G is monotonic decreasing function；D is gradient operator, and Δ is Laplace operator；αi
For weight parameter；GσFor Gaussian kernel；Y0Brightness value for target gray image；To use the obtained chromatic value of different characteristic；
Λ is the function admissible space of the model, and V is any function in function admissible space；
Solve the optimal solution that coupling total variation model obtains pixel t chromatic values, optimal solution using the chromatic value acquired is as repairing
Chromatic value after just, brightness value remain unchanged；
Step 4, repeating said steps 3 are modified each pixel t of target gray image and by revised pixel
The image output of point composition, that is, complete the coloring to target gray image.
2. according to claim 1 based on color transfer and modified image rendering methods, which is characterized in that the step 1
It further includes and selected YCbCr space is normalized.
3. according to claim 1 based on color transfer and modified image rendering methods, which is characterized in that the step 2
Selected pixels point s is specifically included：
The sampling set being made of several pixels is chosen in the coloured image of source；
Calculate in the sampling set pixel the Variance feature of determining window and discrete Fourier transform feature, and choose it
In with pixel t pixel similar in the index value asked under determining window Variance feature and discrete Fourier transform feature make
For feasible solution；
Any one solution is selected from several feasible solutions as target solution, and extracts the pixel coloration corresponding to the target solution
Value.
4. according to claim 1 based on color transfer and modified image rendering methods, which is characterized in that the step 3
Coupling total variation model is solved using ADM algorithms.
5. according to claim 4 based on color transfer and modified image rendering methods, which is characterized in that the ADM is calculated
Method specifically includes：Coupling total variation model is changed into the convex programming problem of the linear convex set contrained of standard；Using ADM algorithms into
Row alternating iteration, which solves, obtains optimal solution.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610112647.2A CN105761292B (en) | 2016-02-29 | 2016-02-29 | One kind is based on color transfer and modified image rendering methods |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610112647.2A CN105761292B (en) | 2016-02-29 | 2016-02-29 | One kind is based on color transfer and modified image rendering methods |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105761292A CN105761292A (en) | 2016-07-13 |
| CN105761292B true CN105761292B (en) | 2018-06-12 |
Family
ID=56330510
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610112647.2A Active CN105761292B (en) | 2016-02-29 | 2016-02-29 | One kind is based on color transfer and modified image rendering methods |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105761292B (en) |
Families Citing this family (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105761284B (en) * | 2016-02-23 | 2017-11-03 | 湖北大学 | A kind of coloured image goes color method and system |
| CN107067443A (en) * | 2017-04-01 | 2017-08-18 | 南京邮电大学 | A kind of brain tumor area coloring method based on MRI medical images |
| CN109903348B (en) * | 2018-05-14 | 2023-11-17 | 秦皇岛知聚科技有限公司 | Interactive image color editing method based on blocking features and digital image processing system |
| CN109816737B (en) * | 2019-01-31 | 2022-08-09 | 南京邮电大学 | Edge preserving image coloring method based on coupling natural vector total variation |
| CN111612870B (en) * | 2019-02-25 | 2023-06-20 | 阿里巴巴集团控股有限公司 | Image processing method and device |
| CN112073596B (en) * | 2020-09-18 | 2021-07-20 | 青岛大学 | Simulated color processing method and system for specific black-and-white video signal |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1604142A (en) * | 2003-07-14 | 2005-04-06 | 三星电子株式会社 | Image-based rendering and editing method and apparatus |
| CN1942922A (en) * | 2003-04-07 | 2007-04-04 | 克雷沃耶提实验室有限公司 | Image data set with embedded pre-subpixel rendered image |
| CN101477695A (en) * | 2009-01-05 | 2009-07-08 | 东南大学 | Method for fast painting image and video |
| CN102214364A (en) * | 2011-04-27 | 2011-10-12 | 天津大学 | Automatic coloring method of gray level images in combination with histogram regression and texture analysis |
-
2016
- 2016-02-29 CN CN201610112647.2A patent/CN105761292B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1942922A (en) * | 2003-04-07 | 2007-04-04 | 克雷沃耶提实验室有限公司 | Image data set with embedded pre-subpixel rendered image |
| CN1604142A (en) * | 2003-07-14 | 2005-04-06 | 三星电子株式会社 | Image-based rendering and editing method and apparatus |
| CN101477695A (en) * | 2009-01-05 | 2009-07-08 | 东南大学 | Method for fast painting image and video |
| CN102214364A (en) * | 2011-04-27 | 2011-10-12 | 天津大学 | Automatic coloring method of gray level images in combination with histogram regression and texture analysis |
Non-Patent Citations (5)
| Title |
|---|
| Variational models for image colorization via chromaticity and brightness decomposition;Sung Ha Kang 等;《IEEE Transactions on Image Processing》;20071231;第16卷(第9期);第1-26页 * |
| 保色调的黑白卡通图像着色方法;朱薇 等;《计算机辅助设计与图形学学报》;20110331;第23卷(第3期);第392-398页 * |
| 基于多种易用标记的图像着色;张可为 等;《光电子·激光》;20140228;第25卷(第2期);第386-392页 * |
| 基于色彩传递与扩展的图像着色算法;朱黎博 等;《中国图象图形学报》;20100228;第15卷(第2期);第200-205页 * |
| 局部稀疏编码的自然灰度图像着色方法;郝凯 等;《计算机辅助设计与图形学学报》;20110831;第23卷(第8期);第1401-1408页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105761292A (en) | 2016-07-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105761292B (en) | One kind is based on color transfer and modified image rendering methods | |
| CN108830912A (en) | A kind of interactive grayscale image color method of depth characteristic confrontation type study | |
| CN107590786A (en) | A kind of image enchancing method based on confrontation learning network | |
| CN103914699A (en) | Automatic lip gloss image enhancement method based on color space | |
| Burgeth et al. | Morphology for color images via Loewner order for matrix fields | |
| CN105913407B (en) | A method of poly focal power image co-registration is optimized based on differential chart | |
| CN105118076B (en) | Based on over-segmentation and the local image colorization method with global coherency | |
| CN104915975B (en) | A kind of image processing method and system of simulation wax crayon colored drawing | |
| CN107481206A (en) | MIcrosope image background equalization Processing Algorithm | |
| CN112991371B (en) | Automatic image coloring method and system based on coloring overflow constraint | |
| CN106971380A (en) | A kind of contrast enhancing and application of the visual saliency optimization method in golf course figure | |
| CN110322530A (en) | It is a kind of based on depth residual error network can interaction figure picture coloring | |
| US5828819A (en) | Apparatus and method for automatically creating a picture in a style having painting-like coloring using an image processing | |
| Mishra | Contrast limited adaptive histogram equalization (CLAHE) approach for enhancement of the microstructures of friction stir welded joints | |
| CN105825531B (en) | A kind of colouring method and device of game object | |
| Maslennikova et al. | Interactive local color transfer between images | |
| CN107742280A (en) | A kind of image sharpening method and device | |
| CN104851074B (en) | The non local neighborhood gray level image colorization method of feature based similitude | |
| CN104715454A (en) | Anti-aliasing graph overlapping algorithm | |
| CN106780379B (en) | A kind of microscopical colour-image reinforcing method of metering | |
| CN107464213A (en) | The heavy color method of Lab space mapping based on monochromatic colour disk | |
| CN105335980B (en) | A kind of coloured image of suitable image SIFT feature matching turns luminance picture method | |
| Jimenez-Arredondo et al. | Multilevel color transfer on images for providing an artistic sight of the world | |
| CN103295250A (en) | Image colorization method based on L1 mixed norm solving | |
| Sharma et al. | Contrast enhancement of an image using fuzzy logic |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |