CN102542522B - Color separating method and system - Google Patents
Color separating method and system Download PDFInfo
- Publication number
- CN102542522B CN102542522B CN201010620858XA CN201010620858A CN102542522B CN 102542522 B CN102542522 B CN 102542522B CN 201010620858X A CN201010620858X A CN 201010620858XA CN 201010620858 A CN201010620858 A CN 201010620858A CN 102542522 B CN102542522 B CN 102542522B
- Authority
- CN
- China
- Prior art keywords
- color
- energy
- separated
- pixel
- category
- 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
Images
Abstract
The invention provides a color separating method and system, which aim at solving the problem of lower color separation efficiency in the prior art. According to the invention, the method comprises the following steps of: determining color type and color mixing rule used during separating colors of an image; establishing a color separation energy model E=f(E1, E2, E3, E4), wherein E represents color separation energy, f represents a preset energy model function, E1 is similar energy which represents similarity of pixel point color and color after the colors are separated, E2 is color mixing energy which represents energy component determined according to the number of the pixel point color types after the colors are separated, E3 is edge energy which represents the feasibility of showing adjacent pixel points in different colors, and E4 is transitional energy which represents the transition feasibility of different colors; carrying out optimization computation on the color separation energy according to the energy model; and using the color after the colors of each pixel are separated when the color separation energy is minimum as the color separation result of the image.
Description
Technical field
The present invention relates to a kind of method and system of color-separated.
Background technology
Color-separated is image processing techniques before a kind of common seal, its objective is a width coloured image is divided into to a plurality of monochrome images according to selected color, and the effect after these monochrome image colour mixtures is as far as possible consistent with former figure.In current printing industry, color-separated is manually carried out in main dependence, thereby efficiency is lower.
For the lower problem of color-separated efficiency of the prior art, effective solution is proposed not yet at present.
Summary of the invention
The present invention aims to provide a kind of method and system of color-separated, to solve the lower problem of color-separated efficiency in prior art.
To achieve these goals, according to an aspect of the present invention, provide a kind of method of color-separated.
The method of color-separated of the present invention comprises: step 1: color category and the colour mixture rule of determining use when image is carried out to color-separated; Step 2: set up color-separated energy model E=f (E
1, E
2, E
3, E
4), wherein E means the color-separated energy, f means the energy model of presetting function, E
1For similar energy, the similarity of color after expression pixel color and color-separated, E
2For the colour mixture energy, mean the energy component of determining according to the number of pixel color category after color-separated, E
3For edge energy, mean neighbor pixel is expressed as the feasibility of different colours, E
4For the transition energy, mean the feasibility of carrying out transition between different colours; Step 3: according to described energy model, be optimized calculating for described color-separated energy, using described color-separated energy after the color-separated of hour each pixel color as the color-separated result of described image.
Further, the similarity of described color adopts the distance of color vector to measure.
Further, in described step 3, the similarity of described color adopts following steps to calculate: the point that will be evenly distributed on the residing color space some of color category is considered as the known color point; Calculate the concentration of each color category that known color point is corresponding with described color category minimum difference degree; When calculating pixel color and described color category similarity, according to the concentration data with described color several nearer known color points in color space, carry out interpolation and obtain the concentration of mating with described pixel color, according to densimeter, calculate the colour mixture color again, finally adopt the distance of color vector to calculate the similarity of color.
Further, described transition energy is the number of the color category of use before and after color-separated.
Further, described energy model is E=α E
1+ β E
2+ μ E
3E
4Or E=α E
1E
2+ β E
3E
4, wherein α, β and μ are default weighting coefficient.
Further, described energy model is E=α E
1+ β E
2+ μ E
3+ vE
4Or E=α E
1E
2+ β E
3+ μ E
4, wherein α, β, μ and v are default weighting coefficient.
Further, before described step 1, also comprise: the thumbnail that generates described image; Described step 2 and step 3 are based on described thumbnail and carry out; And after described step 3, also comprise: the color-separated result according to described thumbnail is carried out color-separated to described image.
Further, according to the color-separated result of described thumbnail, described image being carried out to color-separated comprises: described image is divided into to a plurality of image blocks; For each image block in described a plurality of image blocks, carry out successively described step 2 and step 3, wherein when execution step, according to the color-separated result of thumbnail, calculate the similar energy that each pixel may color category 3 the time, using the color-separated result of optimum solution as described pixel.
Further, calculating each pixel may also comprise before the similar energy of color category: according to the scale down of described thumbnail, determine the subpixel coordinates of each pixel correspondence in thumbnail in described image; For each pixel in described image, the union of the subpixel coordinates that this pixel is corresponding color category of a plurality of pixels in default neighborhood in described thumbnail may color category as the described of this pixel.
Further, described optimization is calculated and is comprised the steps: the first step: artificially arrange or produce at random an initial solution x0, making xbest=x0, and calculate energy function value E (x0) corresponding to this solution; Second step: initial temperature T (0)=T0 is set, and wherein function T (t) is cooling program, and expression formula is T (t)=T0/log (1+t), and wherein T0 is constant, establishes iterations i=1, j=1; The 3rd step: current optimum solution xbest, according to a certain neighborhood function, is produced to a new solution xnew, calculate new energy function value E (xnew), and the increment Delta E=E of calculating energy functional value (xnew)-E (xbest); The 4th step: if Δ E<0, xbest=xnew; If Δ E>0, p=exp (Δ E/T (i)), if the random number c produced between 0 to 1 is less than p, xbest=xnew, otherwise xbest=xbest; The 5th step: j=j+1, if j<K returns to the 3rd step, otherwise carries out next step; Wherein K is default constant; The 6th step: i=i+1, if T (i)>Tmin returns to the 3rd step, otherwise carries out next step; Wherein Tmin is default constant; The 7th step: export current optimum solution.
A kind of system of the color-separated towards spot color is provided according to a further aspect in the invention.
The system of color-separated of the present invention comprises: preserve module, color category and the colour mixture rule of use while for preserving, image being carried out to color-separated; The energy model module, be used to setting up color-separated energy model E=f (E
1, E
2, E
3, E
4), wherein E means the color-separated energy, f means the energy model of presetting function, E
1For similar energy, the similarity of color after expression pixel color and color-separated, E
2For the colour mixture energy, mean the energy component of determining according to the number of pixel color category after color-separated, E
3For edge energy, mean neighbor pixel is expressed as the feasibility of different colours, E
4For the transition energy, mean the feasibility of carrying out transition between different colours; The Optimization Solution module, for according to described energy model, be optimized calculating for described color-separated energy, using described color-separated energy after the color-separated of hour each pixel color as the color-separated result of described image.
Further, described Optimization Solution module is also for calculating in accordance with the following steps the similarity of described color: the point that will be evenly distributed on the residing color space some of color category is considered as the known color point; Calculate the concentration of each color category that known color point is corresponding with described color category minimum difference degree; When calculating pixel color and described color category similarity, according to the concentration data with described color several nearer known color points in color space, carry out interpolation and obtain the concentration of mating with described pixel color, according to densimeter, calculate the colour mixture color again, finally adopt the distance of color vector to calculate the similarity of color.
Further, described energy model module is also be used to setting up the energy model of following form: E=α E
1+ β E
2+ μ E
3E
4, or set up the energy model of following form: E=α E
1E
2+ β E
3E
4, wherein α, β and μ are default weighting coefficient.
Further, described energy model module is also be used to setting up the energy model of following form: E=α E
1+ β E
2+ μ E
3+ vE
4, or set up the energy model of following form: E=α E
1E
2+ β E
3+ μ E
4, wherein α, β, μ and v are default weighting coefficient.
Further, described system also comprises: the thumbnail module, be used to generating the thumbnail of described image; The color-separated module, carry out color-separated for the color-separated result according to described thumbnail to described image; And described energy model module is also for setting up the color-separated energy model based on described thumbnail.
Further, described color-separated module also for: described image is divided into to a plurality of image blocks; For each image block in described a plurality of image blocks, call successively described energy model module and Optimization Solution module, wherein the Optimization Solution module calculates according to the color-separated result of thumbnail the similar energy that each pixel may color category when calculating, and usings the color-separated result of optimum solution as described pixel.
Further, described color-separated module also for: according to the scale down of described thumbnail, determine each pixel in described image corresponding subpixel coordinates in thumbnail; For each pixel in described image, the union of the subpixel coordinates that this pixel is corresponding color category of a plurality of pixels in default neighborhood in described thumbnail may color category as the described of this pixel.
Further, described Optimization Solution module is also for calculating in accordance with the following steps: the first step: artificially arrange or produce at random an initial solution x0, making xbest=x0, and calculate energy function value E (x0) corresponding to this solution; Second step: initial temperature T (0)=T0 is set, and wherein function T (t) is cooling program, and expression formula is T (t)=T0/log (1+t), and wherein T0 is constant, establishes iterations i=1, j=1; The 3rd step: current optimum solution xbest, according to a certain neighborhood function, is produced to a new solution xnew, calculate new energy function value E (xnew), and the increment Delta E=E of calculating energy functional value (xnew)-E (xbest); The 4th step: if Δ E<0, xbest=xnew; If Δ E>0, p=exp (Δ E/T (i)), if the random number c produced between 0 to 1 is less than p, xbest=xnew, otherwise xbest=xbest; The 5th step: j=j+1, if j<K returns to the 3rd step, otherwise carries out next step; Wherein K is default constant; The 6th step: i=i+1, if T (i)>Tmin returns to the 3rd step, otherwise carries out next step; Wherein Tmin is default constant; The 7th step: export current optimum solution.
Apply technical scheme of the present invention, by the color-separated energy model to including similar energy, colour mixture energy, transition energy and edge energy, be optimized and solve, can obtain higher color-separated work efficiency, in general, color for 3～4 color-separated uses, the color separation used time only needs a few minutes, and identical image is carried out to manual color separation needs several hours even several days.The technical scheme of the embodiment of the present invention can effectively be increased work efficiency as can be seen here.
The accompanying drawing explanation
Figure of description is used to provide a further understanding of the present invention, forms the application's a part, and schematic description and description of the present invention the present invention does not form inappropriate limitation of the present invention for explaining.In the accompanying drawings:
Fig. 1 is the key step schematic diagram according to the method for the color-separated of the embodiment of the present invention;
Fig. 2 is the schematic diagram according to the color-separated based on thumbnail of the embodiment of the present invention;
Fig. 3 is the schematic diagram according to the main modular of the system of the color-separated of the embodiment of the present invention; And
Fig. 4 A to Fig. 4 E is the schematic diagram according to the effect of the color-separated of the embodiment of the present invention.
Embodiment
It should be noted that, in the situation that do not conflict, embodiment and the feature in embodiment in the application can combine mutually.Describe below with reference to the accompanying drawings and in conjunction with the embodiments the present invention in detail.
Fig. 1 is the key step schematic diagram according to the method for the color-separated of the embodiment of the present invention, and as shown in Figure 1, the method mainly comprises the steps:
Step 1: color category and the colour mixture rule of determining use when image is carried out to color-separated.Image in this step is the image of the processing of pending color allocation.
Step 2: set up color-separated energy model E=f (E
1, E
2, E
3, E
4), wherein E means the color-separated energy, f means the energy model of presetting function, E
1For similar energy, the similarity of color after expression pixel color and color-separated, E
2For the colour mixture energy, mean the energy component of determining according to the number of pixel color category after color-separated, E
3For edge energy, mean neighbor pixel is expressed as the feasibility of different colours, E
4For the transition energy, mean the feasibility of carrying out transition between different colours.
Step 3: according to energy model, be optimized calculating for described color-separated energy, using the color-separated energy after the color-separated of hour each pixel color as the color-separated result of described image.
In above-mentioned steps, the similarity of color can adopt the distance of color vector to measure, and the transition energy can be the number of the color category of use before and after color-separated.
In step 3, the similarity of color specifically can adopt following steps to calculate:
The point that is evenly distributed on the residing color space some of color category is considered as to the known color point;
Calculate the concentration of each color category that known color point is corresponding with described color category minimum difference degree;
When calculating pixel color and described color category similarity, according to the concentration data with described color several nearer known color points in color space, carry out interpolation and obtain the concentration of mating with described pixel color, according to densimeter, calculate the colour mixture color again, finally adopt the distance of color vector to calculate the similarity of color.
Energy model can be specifically E=α E
1+ β E
2+ μ E
3E
4Or E=α E
1E
2+ β E
3E
4, wherein α, β and μ are default weighting coefficient; Energy model can be also E=α E
1+ β E
2+ μ E
3+ vE
4Or E=α E
1E
2+ β E
3+ μ E
4, wherein α, β, μ and v are default weighting coefficient.
In the present embodiment, can adopt thumbnail to carry out color-separated, specifically before step 1, generate the thumbnail of pending image, like this, step 2 and step 3 are carried out based on this thumbnail.And the color-separated result according to thumbnail after step 3 is carried out color-separated to image, below this is illustrated.
When the color-separated result according to thumbnail is carried out color-separated to image, can first image be divided into to a plurality of image blocks; Then for each image block in a plurality of image blocks, perform step successively 2 and step 3, wherein when execution step, according to the color-separated result of thumbnail, only calculate the similar energy E 1 that each pixel may color category 3 the time, using the optimum solution of energy E as the color-separated result of pixel.The possible color category here can be determined as follows: at first according to the scale down of thumbnail, determine the subpixel coordinates of each pixel correspondence in thumbnail in image; Then for each pixel in image, the union of the subpixel coordinates that this pixel is corresponding color category of a plurality of pixels in default neighborhood in thumbnail is as the possible color category of this pixel.
In the optimization of step 3 is calculated, can adopt simulated annealing, concrete steps are as follows:
The first step: produce at random an initial solution x0, make xbest=x0, and calculate energy function value E (x0) corresponding to this solution;
Second step: initial temperature T (0)=T0 is set, and wherein function T (t) is cooling program, and expression formula is T (t)=T0/log (1+t), and wherein T0 is constant, establishes iterations i=1, j=1;
The 3rd step: current optimum solution xbest, according to a certain neighborhood function, is produced to a new solution xnew, calculate new energy function value E (xnew), and the increment Delta E=E of calculating energy functional value (xnew)-E (xbest);
The 4th step: if Δ E<0, xbest=xnew; If Δ E>0, p=exp (Δ E/T (i)), if the random number c produced between 0 to 1 is less than p, xbest=xnew, otherwise xbest=xbest;
The 5th step: j=j+1, if j<K returns to the 3rd step, otherwise carries out next step; Wherein K is default constant;
The 6th step: i=i+1, if T (i)>Tmin returns to the 3rd step, otherwise carries out next step; Wherein Tmin is default constant;
The 7th step: export current optimum solution
Below by a concrete example, further illustrate the technical scheme of the embodiment of the present invention.Color-separated in this example (being designated hereinafter simply as " color separation ") mainly 1 to 7 is carried out according to the following steps.
1. generation thumbnail.When if pending picture size large (as: 20,000 * 20,000 pixel) can't disposablely be written into internal memory by all images data, first generate the thumbnail image of a former figure.Concrete generating mode has a lot, as nearest neighbor method, and linear interpolation method etc.Thumbnail size is unfixing, depending on the calculator memory situation, generally is no more than 1024 * 1024 pixels.
2. determine the color category that color separation is used.This step needs artificial color and the colour mixture rule of carrying out color separation of setting.The colour mixture rule is used for limiting between each color separation color the situation that colour mixture occurs, and specific implementation method comprises: maximum colour mixture number is set, according to institute likely the colour mixture situation select, manually input rule, or above-mentioned rule in conjunction with etc.The colour mixture rule can not arrange especially yet, and now the whole colors that participate in color separation of acquiescence all colour mixture can occur arbitrarily.In case determined color separation color and colour mixture rule, all colours kind of using when color separation has also just been determined.For instance, suppose that the color separation color selected three kinds, be respectively a, b, c, if the color separation rule is not set, the color category that so final color separation is used is 7 classes, is respectively: a, b, c, a and b two colour mixtures, a and c two colour mixtures, b and c two colour mixtures, a and b and c tri-colour mixtures; If the color separation rule be set as maximum colour mixture number be 2 and a and c can not colour mixture, the color category that uses of final color separation is 5 classes, a, b, c, a and b two colour mixtures, b and c two colour mixtures.
3. energy calculates.According to the color separation energy model, calculate.The color separation energy model comprises following four: similar energy, colour mixture energy, transition energy, edge energy (concrete title is not unique, but all meets the definition of described cost).Wherein,
3.1 similar energy: weigh the color of certain pixel in image and the similarity of each color category that color separation is used.
In a color category, may have one or more constituents, the variable concentrations collocation of these constituents can produce different blend colors.While calculating similarity, measure with each possible blend color of this color category, the most similar outcome record is got off.The similar energy datum number of final entry equals pixel count * color category number.
Blend color calculates or tables look-up according to concrete color space and actual colour mixture mode.Available following formula calculates blend color in the CMYK space.
Calculating in realization can be not limited to this formula.Also can in advance blend color numerical value be write in form, table look-up when calculating.
Concrete measure can be for pixel color and colour mixture color the distance measure in certain color space, as the Euclidean distance in the Lab space, also can adopt colour difference formula, such as CIE94, the method that can embody in a word two color data differences gets final product.
Can calculate each possible colour mixture color and the similarity of this pixel color in certain color category by the method for exhaustion, but if the concentration rank is large or/and color constituent when more, the method of exhaustion can be very consuming time, and actual application value can greatly reduce even can't be practical.For instance, certain color category is grouped into by 3 kinds of one-tenth, and every kind of composition has 100 concentration ranks, and the colour mixture color relation that this color category comprises has 100 * 100 * 100, supposing to carry out similarity, to calculate the used time be 1 microsecond, calculates that this color category similarity of a pixel is consuming time just reaches 1 second.In prediction on such basis, for the image of 1024 * 1024 pixels, only calculate the similarity of a color category will 1048576 seconds consuming time ≈ 12 days.
Designed a kind of method of quick calculating similarity.The method solves by optimal way, and majorized function is designed to min E (α)=diff (C
Pixel, C
Mixed(α)).C wherein
PixelFor pixel color, C
Mixed(α) for each constituent concentration of this color category, get α (α=[α
1, α
2..., α
n], n is for forming the composition number of this color category) time blend color, diff () means to ask the diversity factor computing.An Euclidean distance that embodiment is two data of diversity factor.An embodiment of optimization method is PSO (particle group optimizing) algorithm.
The another kind of method of calculating fast similarity is to solve by interpolation method.Design philosophy is that each point in certain color category color space of living in is considered as to a known color, use the method for exhaustion or aforementioned optimized algorithm to calculate each constituent concentration that this point is corresponding with this color category minimum difference degree, so calculate enough a plurality of points and record corresponding concentration data.When needing the similarity of calculating pixel color and this color category, can be according to the concentration data with this pixel color nearer several points in color space, carry out interpolation and obtain the concentration of mating with pixel color, according to densimeter, calculate blend color again, obtain similarity by the measure defined.
The present invention is incorporated into enforcement use by above-mentioned two quick calculation methods, and color space used is the CMY space, and counting of calculating is 10 * 10 * 10=1000 and is evenly distributed in space.The interpolation algorithm used is the three dimensions linear interpolation algorithm, and computing method are as follows:
If (fc, fm, fy) is the coordinate of pixel in the CMY space divided by 10 result, α (c, m, y) means the concentration vector that in this space, (c, m, y) locates.Make (nc, nm, ny) and (dc, dm, dy) be respectively integral part and the fraction part of (fc, fm, fy), that is, and dc=fc-nc, dm=fm-nm, dy=fy-ny.{。##.##1},
α(fc，fm，fy)＝(1-dc)·(1-dm)·(1-dy)·α(nc，nm，ny)+dc·(1-dm)·(1-dy)·α(nc+1，nm，ny)+(1-dc)·dm·(1-dy)·α(nc，nm+1，ny)+dc·dm·(1-dy)·α(nc+1，nm+1，ny)+(1-dc)·(1-dm)·dy·α(nc，nm，ny+1)+dc·(1-dm)·dy·α(nc+1，nm，ny+1)+(1-dc)·dm·dy·α(nc，nm+1，ny+1)+dc·dm·dy·α(nc，nm，ny)
3.2 colour mixture energy: weigh certain color category and form the number that comprises color component.
In color separation, need to be controlled the color mixture situation, for these characteristics, designed the colour mixture energy term.When calculating this energy, any relation of considered pixel intrinsic colour and color category not, and only be concerned about the constituent of color category itself.If wish the generation blend of colors situation of trying one's best few in color separation, the color component number of color category is more, and the colour mixture energy values is just larger; Otherwise if wish generation blend of colors as much as possible, the color component number of color category is more, the colour mixture energy values is just less.The specific design mode of this energy has a lot, all can as long as meet foregoing description.An embodiment is: this energy is proportional to the color number that color category comprises.
3.3 edge energy: the similarity of weighing in image color between two neighbors.
This energy is used for measuring two neighbors is divided into to the cost that the different colours kind will be paid.The definition of neighbor can be one or more pixels that distance is close, and is not limited only to the adjacent pixels in image.Design philosophy is, the edge energy when two similar pixels are separated is larger, and the edge energy when two dissimilar pixels are separated is less.The method of weighing similarity can be with reference to the method in a.An embodiment is the inverse of two pixel color Euclidean distances.
3.4 transition energy: weigh the feasibility that is transitioned into another color category from a color category.
This energy term is used for weighing from a color category and is transitioned into the intrinsic cost that another color category will be paid.In actual color separation, usually require between color and color transition as far as possible smooth.Follow this principle, when this energy of design, can consider the different colours kind whether contain the same color composition and contain the number come to determine, the color component number sum that circular is two labels deducts the number of same color in two labels.The value of 3 color color separations is as shown in table 1 below.
Table 1
| Look 1 | Look 2 | |
Look 1+2 | Look 1+3 | Look 2+3 | Look 1+2+3 | |
| Look 1 | 0 | 2 | 2 | 1 | 1 | 3 | 2 |
| Look 2 | 2 | 0 | 2 | 1 | 3 | 1 | 2 |
| |
2 | 2 | 0 | 3 | 1 | 1 | 2 |
| Look 1+2 | 1 | 1 | 3 | 0 | 2 | 2 | 1 |
| Look 1+3 | 1 | 3 | 1 | 2 | 0 | 2 | 1 |
| Look 2+3 | 3 | 1 | 1 | 2 | 2 | 0 | 1 |
| Look 1+2+3 | 2 | 2 | 2 | 1 | 1 | 1 | 0 |
4. energy-optimised.This energy can be used multiple known optimization method to solve.Can adopt simulated annealing above to solve.
5. export the thumbnail result.According to the optimum results in 4, determine the color category of each pixel in image, and record.
6. according to the image block of thumbnail result treatment full figure.With reference to figure 2, Fig. 2 is the schematic diagram according to the color-separated based on thumbnail of the embodiment of the present invention.In figure, 21 mean original image, and 22 mean the thumbnail image of original image, and 23 mean the amplification of the part in thumbnail image.Original image is divided into to a plurality of image blocks, and the size of each piece is less than or equal to once accessible maximum image of calculator memory.While processing each piece, repeating step 3 and 4.During repeating step 3, with reference to the thumbnail result, only solve the similar energy of the color category that each pixel is possible.Concrete grammar is: current point coordinate is tried to achieve to its subpixel coordinates in thumbnail by the scale down in step 1, according to this coordinate position 4 pixels that find it to close on as shown below.The union of the separately color category of these four pixels in the thumbnail result is exactly all possible color category situation of current point.
Certainly, for large figure, also can not use thumbnail, directly piecemeal is done, but may produce inconsistent effect at the boundary of image block.
7. the color separation result of output image piece.Optimum results according to 6, determine the color category of each pixel in image, uses the similarity calculating method in 3 (a), obtains the concentration of each composition, changes into gray scale and write virtual Color Channel corresponding in result images.
Fig. 3 is the schematic diagram according to the main modular of the system of the color-separated of the embodiment of the present invention, and as shown in Figure 3, the system 30 of the color-separated of the embodiment of the present invention mainly comprises as lower module:
Preserve module, color category and the colour mixture rule of use while for preserving, image being carried out to color-separated;
The energy model module, be used to setting up color-separated energy model E=f (E
1, E
2, E
3, E
4), wherein E means the color-separated energy, f means the energy model of presetting function, E
1For similar energy, the similarity of color after expression pixel color and color-separated, E
2For the colour mixture energy, mean the energy component of determining according to the number of pixel color category after color-separated, E
3For edge energy, mean neighbor pixel is expressed as the feasibility of different colours, E
4For the transition energy, mean the feasibility of carrying out transition between different colours;
The Optimization Solution module, for according to described energy model, be optimized calculating for described color-separated energy, using described color-separated energy after the color-separated of hour each pixel color as the color-separated result of described image.
The Optimization Solution module also can be used for calculating in accordance with the following steps the similarity of described color: the point that will be evenly distributed on the residing color space some of color category is considered as the known color point; Calculate the concentration of each color category that known color point is corresponding with described color category minimum difference degree; When calculating pixel color and described color category similarity, according to the concentration data with described color several nearer known color points in color space, carry out interpolation and obtain the concentration of mating with described pixel color, according to densimeter, calculate the colour mixture color again, finally adopt the distance of color vector to calculate the similarity of color.
The energy model module also can be used for setting up the energy model of following form: E=α E
1+ β E
2+ μ E
3E
4, or set up the energy model of following form: E=α E
1E
2+ β E
3E
4, wherein α, β and μ are default weighting coefficient.
The energy model module also can be used for setting up the energy model of following form: E=α E
1+ β E
2+ μ E
3+ vE
4, or set up the energy model of following form: E=α E
1E
2+ β E
3+ μ E
4, wherein α, β, μ and v are default weighting coefficient.
The system 30 of color-separated also can comprise as lower module (illustrating not shown): the thumbnail module, be used to generating the thumbnail of described image; The color-separated module, carry out color-separated for the color-separated result according to described thumbnail to described image; And described energy model module is also for setting up the color-separated energy model based on described thumbnail.
The color-separated module also can be used for: described image is divided into to a plurality of image blocks; For each image block in described a plurality of image blocks, call successively described energy model module and Optimization Solution module, wherein the Optimization Solution module calculates according to the color-separated result of thumbnail the similar energy E 1 that each pixel may color category when calculating, and usings the optimum solution of energy E as the color-separated result of described pixel.
The color-separated module also can be used for: according to the scale down of described thumbnail, determine the subpixel coordinates of each pixel correspondence in thumbnail in described image; For each pixel in described image, the union of the subpixel coordinates that this pixel is corresponding color category of a plurality of pixels in default neighborhood in described thumbnail may color category as the described of this pixel.
The Optimization Solution module also can be used for calculating in accordance with the following steps:
The first step: artificially arrange or produce at random an initial solution x0, making xbest=x0, and calculate energy function value E (x0) corresponding to this solution;
Second step: initial temperature T (0)=T0 is set, and wherein function T (t) is cooling program, and expression formula is T (t)=T0/log (1+t), and wherein T0 is constant, establishes iterations i=1, j=1;
The 3rd step: current optimum solution xbest, according to a certain neighborhood function, is produced to a new solution xnew, calculate new energy function value E (xnew), and the increment Delta E=E of calculating energy functional value (xnew)-E (xbest);
The 4th step: if Δ E<0, xbest=xnew; If Δ E>0, p=exp (Δ E/T (i)), if the random number c produced between 0 to 1 is less than p, xbest=xnew, otherwise xbest=xbest;
The 5th step: j=j+1, if j<K returns to the 3rd step, otherwise carries out next step; Wherein K is default constant;
The 6th step: i=i+1, if T (i)>Tmin returns to the 3rd step, otherwise carries out next step; Wherein Tmin is default constant;
The 7th step: export current optimum solution.
Fig. 4 A to Fig. 4 E is the schematic diagram according to the effect of the color-separated of the embodiment of the present invention.Wherein Fig. 4 A is the image that original image is namely treated color separation, a kind of color that the color of extracting its upper right portion is used during as color-separated, this color is (pressing CMYK tetra-looks): C:13%, M:67%, Y:90%, K:0%, three kinds of colors of other that use during color-separated are respectively C100, M100 and Y100.Fig. 4 B to Fig. 4 E is respectively the result images of the color-separated that these four passages of spot color of color C100, M100 after color separation, Y100 and said extracted are corresponding.
The embodiment of the present invention is optimized and solves by the color-separated energy model to including similar energy, colour mixture energy, transition energy and edge energy, can obtain higher color-separated work efficiency, in general, color for 3～4 color-separated uses, the color separation used time only needs a few minutes, and identical image is carried out to manual color separation needs several hours even several days.The technical scheme of the embodiment of the present invention can effectively be increased work efficiency as can be seen here.
The foregoing is only the preferred embodiments of the present invention, be not limited to the present invention, for a person skilled in the art, the present invention can have various modifications and variations.Within the spirit and principles in the present invention all, any modification of doing, be equal to replacement, improvement etc., within all should being included in protection scope of the present invention.
Claims (18)
1. the method for a color-separated, is characterized in that, comprising:
Step l: color category and the colour mixture rule of determining use when image is carried out to color-separated;
Step 2: set up color-separated energy model E=f (E
1, E
2, E
3, E
4), wherein E means the color-separated energy, f means the energy model of presetting function, E
1For similar energy, the similarity of color after expression pixel color and color-separated, E
2For the colour mixture energy, mean the energy component of determining according to the number of pixel color category after color-separated, E
3For edge energy, mean neighbor pixel is expressed as the feasibility of different colours, E
4For the transition energy, mean the feasibility of carrying out transition of asking of different colours;
Step 3: according to described energy model, be optimized calculating for described color-separated energy, using described color-separated energy after the color-separated of hour each pixel color as the color-separated result of described image.
2. method according to claim 1, is characterized in that, the similarity of described color adopts the distance of color vector to measure.
3. method according to claim 1 and 2, is characterized in that, in described step 3, the similarity of described color adopts following steps to calculate:
By being evenly distributed on the residing color sky of color category, ask that the point of some is considered as the known color point;
Calculate the concentration of each color category that known color point is corresponding with described color category minimum difference degree;
When calculating pixel color and described color category similarity, according to the concentration data with described color several nearer known color points in the color sky is asked, carry out interpolation and obtain the concentration of mating with described pixel color, according to densimeter, calculate the colour mixture color again, finally adopt the distance of color vector to calculate the similarity of color.
4. method according to claim 1, is characterized in that, described transition energy is the number of the color category of use before and after color-separated.
5. method according to claim 1, is characterized in that, described energy model is E=α E
1+ β E
2+ μ E
3E
4Or E=α E
1E
2+ β E
3E
4, wherein α, β and μ are default weighting coefficient.
6. method according to claim 1, is characterized in that, described energy model is E=α E
1+ β E
2+ μ E
3+ ν E
4Or E=α E
1E
2+ β E
3+ μ E
4, wherein α, β, μ and ν are default weighting coefficient.
7. method according to claim 1, is characterized in that, also comprises:
Before described step l, also comprise: the thumbnail that generates described image;
Described step 2 and step 3 are based on described thumbnail and carry out; And
After described step 3, also comprise: the color-separated result according to described thumbnail is carried out color-separated to described image.
8. method according to claim 7, is characterized in that, according to the color-separated result of described thumbnail, described image carried out to color-separated and comprise:
Described image is divided into to a plurality of image blocks;
For each image block in described a plurality of image blocks, carry out successively described step 2 and step 3, wherein when execution step, according to the color-separated result of thumbnail, calculate the similar energy that each pixel may color category 3 the time, using the color-separated result of optimum solution as described pixel.
9. method according to claim 8, is characterized in that, calculating each pixel may also comprise before the similar energy of color category:
According to the scale down of described thumbnail, determine the subpixel coordinates of each pixel correspondence in thumbnail in described image;
For each pixel in described image, the union of the subpixel coordinates that this pixel is corresponding color category of a plurality of pixels in default neighborhood in described thumbnail may color category as the described of this pixel.
10. method according to claim 1, is characterized in that, described optimization is calculated and comprised the steps:
The first step: artificially arrange or produce at random an initial solution x0, making xbest=x0, and calculate energy function value E (x0) corresponding to this initial solution;
Second step: initial temperature T (0)=T0 is set, and wherein function T (t) is cooling program, and expression formula is T (t)=T0/log (1+t), and wherein T0 is constant, establishes iterations i=1, j=1;
The 3rd step: current optimum solution xbest, according to a certain neighborhood function, is produced to a new solution xnew, calculate new energy function value E (xnew), and the increment △ E=E (xnew) of calculating energy functional value-E (xbest);
The 4th step: if △ is E<0, xbest=xnew; If △ is E > 0, p=exp (△ E/T (i)), if the random number c asked produced 0 to 1 is less than p, xbest=xnew, otherwise xbest=xbest;
The 5th step: j=j+1, if j<K returns to the 3rd step, otherwise carries out next step; Wherein K is default constant;
The 6th step: i=i+1, if T (i) > Tmin, return to the 3rd step, otherwise carry out next step; Wherein Tmin is default constant;
The 7th step: export current optimum solution.
11. the system of a color-separated, is characterized in that, comprising:
Preserve module, color category and the colour mixture rule of use while for preserving, image being carried out to color-separated;
The energy model module, be used to setting up color-separated energy model E=f (E
1, E
2, E
3, E
4), wherein E means the color-separated energy, f means the energy model of presetting function, E
1For similar energy, the similarity of color after expression pixel color and color-separated, E
2For the colour mixture energy, mean the energy component of determining according to the number of pixel color category after color-separated, E
3For edge energy, mean neighbor pixel is expressed as the feasibility of different colours, E
4For the transition energy, mean the feasibility of carrying out transition of asking of different colours;
The Optimization Solution module, for according to described energy model, be optimized calculating for described color-separated energy, using described color-separated energy after the color-separated of hour each pixel color as the color-separated result of described image.
12. system according to claim 11, is characterized in that, described Optimization Solution module is also for calculating in accordance with the following steps the similarity of described color:
By being evenly distributed on the residing color sky of color category, ask that the point of some is considered as the known color point;
Calculate the concentration of each color category that known color point is corresponding with described color category minimum difference degree;
When calculating pixel color and described color category similarity, according to the concentration data with described color several nearer known color points in the color sky is asked, carry out interpolation and obtain the concentration of mating with described pixel color, according to densimeter, calculate the colour mixture color again, finally adopt the distance of color vector to calculate the similarity of color.
13. system according to claim 11, is characterized in that,
Described energy model module is also be used to setting up the energy model of following form: E=α E
1+ β E
2+ μ E
3E
4Or E=α E
1E
2+ β E
3E
4, wherein α, β and μ are default weighting coefficient.
14. system according to claim 11, is characterized in that,
Described energy model module is also be used to setting up the energy model of following form: E=α E
1+ β E
2+ μ E
3+ ν E
4Or E=α E
1E
2+ β E
3+ μ E
4, wherein α, β, μ and ν are default weighting coefficient.
15. system according to claim 11, is characterized in that, also comprises:
The thumbnail module, be used to generating the thumbnail of described image;
The color-separated module, carry out color-separated for the color-separated result according to described thumbnail to described image;
And described energy model module is also for setting up the color-separated energy model based on described thumbnail.
16. system according to claim 11, is characterized in that, described color-separated module also for:
Described image is divided into to a plurality of image blocks;
For each image block in described a plurality of image blocks, call successively described energy model module and Optimization Solution module, wherein the Optimization Solution module calculates according to the color-separated result of thumbnail the similar energy that each pixel may color category when calculating, and usings the color-separated result of optimum solution as described pixel.
17. system according to claim 16, is characterized in that, described color-separated module also for:
According to the scale down of described thumbnail, determine the subpixel coordinates of each pixel correspondence in thumbnail in described image;
For each pixel in described image, the union of the subpixel coordinates that this pixel is corresponding color category of a plurality of pixels in default neighborhood in described thumbnail may color category as the described of this pixel.
18. system according to claim 11, is characterized in that, described Optimization Solution module is also for calculating in accordance with the following steps:
The first step: artificially arrange or produce at random an initial solution x0, making xbest=x0, and calculate energy function value E (x0) corresponding to this initial solution;
Second step: initial temperature T (0)=T0 is set, and wherein function T (t) is cooling program, and expression formula is T (t)=T0/log (1+t), and wherein T0 is constant, establishes iterations i=l, j=l;
The 3rd step: current optimum solution xbest, according to a certain neighborhood function, is produced to a new solution xnew, calculate new energy function value E (xnew), and the increment △ E=E (xnew) of calculating energy functional value-E (xbest);
The 4th step: if △ is E<0, xbest=xnew; If △ is E > 0, P=exp (△ E/T (i)), if the random number c asked produced 0 to 1 is less than p, xbest=xnew, otherwise xbest=xbest;
The 5th step: j=j+1, if j<K returns to the 3rd step, otherwise carries out next step; Wherein K is default constant;
The 6th step: i=i+1, if T (i) > Tmin, return to the 3rd step, otherwise carry out next step; Wherein Tmin is default constant;
The 7th step: export current optimum solution.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201010620858XA CN102542522B (en) | 2010-12-31 | 2010-12-31 | Color separating method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201010620858XA CN102542522B (en) | 2010-12-31 | 2010-12-31 | Color separating method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102542522A CN102542522A (en) | 2012-07-04 |
| CN102542522B true CN102542522B (en) | 2013-11-27 |
Family
ID=46349348
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201010620858XA Active CN102542522B (en) | 2010-12-31 | 2010-12-31 | Color separating method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102542522B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106251302A (en) * | 2016-07-27 | 2016-12-21 | 合肥高晶光电科技有限公司 | A kind of image processing method based on simulated annealing |
| CN110837351B (en) * | 2019-11-27 | 2023-09-08 | 深圳市汉森软件有限公司 | Pure color processing method, device, equipment and storage medium based on ICC (inter-color computer) processing |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101277370A (en) * | 2007-03-26 | 2008-10-01 | 村田机械株式会社 | Image processing apparatus and method for detecting a background color |
| CN101395899A (en) * | 2006-03-03 | 2009-03-25 | 爱克发印艺公司 | Multicolorant separation system and method |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8131067B2 (en) * | 2008-09-11 | 2012-03-06 | Seiko Epson Corporation | Image processing apparatus, image processing method, and computer-readable media for attaining image processing |
-
2010
- 2010-12-31 CN CN201010620858XA patent/CN102542522B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101395899A (en) * | 2006-03-03 | 2009-03-25 | 爱克发印艺公司 | Multicolorant separation system and method |
| CN101277370A (en) * | 2007-03-26 | 2008-10-01 | 村田机械株式会社 | Image processing apparatus and method for detecting a background color |
Non-Patent Citations (2)
| Title |
|---|
| 史久根 等.彩色地图图象中道路信息的识别和提取.《微机发展》.2000,第10卷(第1期),全文. |
| 彩色地图图象中道路信息的识别和提取;史久根 等;《微机发展》;20000131;第10卷(第1期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102542522A (en) | 2012-07-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108830912B (en) | Interactive gray image coloring method for depth feature-based antagonistic learning | |
| Gooch et al. | Color2gray: salience-preserving color removal | |
| CN102761766B (en) | Method for depth map generation | |
| CN102903128B (en) | The video image content editor's transmission method kept based on Similarity of Local Characteristic Structure | |
| US10668666B2 (en) | Color mapping in 3D printing | |
| CN106485720A (en) | Image processing method and device | |
| CN105761202B (en) | A kind of color image color moving method | |
| CN103793888B (en) | Image enhancing method based on main colors of reference image | |
| CN103617596A (en) | Image color style transformation method based on flow pattern transition | |
| CN105893649B (en) | A kind of interactive image based on optimal model restains method | |
| CN101652988A (en) | Color processor, color processing method, and recording medium | |
| JP2012520045A (en) | Multi-primary conversion | |
| CN103957343B (en) | A kind of colour errors method of diffusion | |
| Milić et al. | A content-dependent naturalness-preserving daltonization method for dichromatic and anomalous trichromatic color vision deficiencies | |
| CN1604657B (en) | Method and apparatus for adjusting colors of an image | |
| CN102542522B (en) | Color separating method and system | |
| CN113436284A (en) | Image processing method and device, computer equipment and storage medium | |
| CN104796682A (en) | Image signal color enhancement method and image signal color enhancement device | |
| US11057545B2 (en) | Weighting functions which vary with intended area color values | |
| US8780409B2 (en) | Method and system for optimizing black point compensation parameters | |
| CN114494467A (en) | Image color migration method and device, electronic equipment and storage medium | |
| CN113012079A (en) | Low-brightness vehicle bottom image enhancement method and device and storage medium | |
| Chen et al. | Optimizing color transfer using color similarity measurement | |
| CN107886550B (en) | Image editing propagation method and system | |
| KR0151918B1 (en) | Image generation apparatus and method for image processing |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |