JP2002042105A - Computer color matching method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a computer color matching method capable of obtaining a more practical combination of color materials and its blending ratio. SOLUTION: In computer color matching, a bit layout indicating a combination of a plurality of color materials at a blending ratio in binary digits is used as a gene, and an evaluation index which is the function of at least one characteristic quantity indicating the difference in color is set. The evaluation index is used to evaluate the genes in a gene group of a plurality of genes, and the gene getting the highest evaluation is preserved by this method. When it is judged that the optimum blending is not obtained, the process includes a step that a higher probability is given to the genes getting higher evaluation in the gene group to select the genes to be retained, a step for interlacing the retained genes, and a step for achieving the mutation of the genes. The above evaluation is made on the obtained gene group, and the above process is repeated until the optimum blending is obtained.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

[0001] 1. Field of the Invention [0002] The present invention relates to computer color matching.

[0002]

2. Description of the Related Art Color matching (toning) involves selecting a color material (dye, etc.) to be used to make a fiber product or the like a target color, and a blending ratio (blending formula) of each color material in the color material combination. Is what you want. In computer color matching, this is calculated by a computer. Usually, in the case of isometric matching, the evaluation as to whether or not the target color matches the target color is made by a function value obtained from the difference between the spectral reflectance of the target color and the spectral reflectance of the color material mixture calculated by a computer (for example, , The sum of the squared values of the reflectance differences of the respective wavelengths, or the value obtained by weighting and summing the reflectance differences of the respective wavelengths). Determine the mixing ratio. In the case of metamerism (equal color matching), the color difference between the target color and the color of the color material mixture calculated by the computer under the designated light source is evaluated, and the color material set is minimized. Determine the combination and mix ratio.

As a general conventional calculation method for finding an optimal solution used in computer color matching,
Some methods use a method (least-squares method) of calculating a mixing ratio of a coloring material so as to minimize the square of an error between the reflectance of a target color and the reflectance obtained by calculation. Further, there is a method of changing the mixing ratio of the coloring material so that the difference Δt between the target color value and the color value at a certain mixing ratio is set to 0 using the following color correction equation. Δt = B ・ ΔC

[Formula 1]

[Formula 2]

[Equation 3] Here, X, Y, Z are tristimulus values of the color, △ t is the change of color value, △ C compounding ratio C ₁ of the three color material,
This is the change amount of C ₂ and C ₃ .

[0004]

In the conventional computer color matching, calculation is performed by either isometric matching or metamerism matching, or
After calculating by isomeric matching, the composition was adjusted by metamerism matching. However, isomeric matching can be used only in a very limited system in which the color material used for the target color and the color material on hand are exactly the same. In a system in which a color material used for a general purpose color is different from a hand-held color material, it is not always possible to obtain a primary color mixture desired by a toner. Also, metamerism matching is that if multiple similar color materials are used for the target color, the composition balance will be extremely poor, and the color difference will match under the specified light source, but the metamerism will increase. is there. For example, when there is no color material used for the target color among the color materials on hand, or when the condition-equivalent color cannot be obtained or when the color match cannot be performed, it is necessary to determine the combination and the mixing ratio of the color material that becomes the closest color. In such a case, in the conventional computer color matching that considers only the color difference, a calculation result different from the sense of the toning person is often output. Also, conventional computer color matching usually involves
Used to determine the composition of the three colorants,
When two color materials were used, a highly accurate result was not obtained, and the reproducibility was poor.

It is an object of the present invention to provide a computer color matching method which requires more practical combinations of color materials and their compounding ratios.

[0006]

According to the computer color matching method of the present invention, at least one characteristic representing a color difference is represented by using a bit array, which represents a combination of a plurality of color material mixing ratios in a binary number, as a gene. Set an evaluation index that is a function of the amount, for a gene population consisting of a plurality of genes, evaluate the genes in the gene population using the evaluation index, and save the best-evaluated genes; and If it is determined that the gene has not been obtained, then, in the gene population, the higher the evaluation of the gene, the higher the probability, the step of selecting the gene to be left, and then, the crossover of the remaining gene And further performing a mutation of the gene, returning to the step of performing the above-described evaluation on the gene population thus obtained, Until engagement is obtained repeating the process described above. Preferably, the function of the characteristic amount representing the color difference includes at least one of a function of a spectral reflectance difference, a metamerism function, and a color function. Preferably, the characteristic value representing the color difference further includes a flop value. Preferably, the evaluation index is obtained by weighting and adding at least two of a function of a spectral reflectance difference, a metamerism function, and a color function. Preferably, the above-mentioned evaluation index is determined in consideration of a contribution rate of each angle based on a value obtained for each of a plurality of angles. Preferably, the initial gene population includes a 100% composition of each colorant and a composition obtained by an initial value calculation based on spectral reflectance data.

[0007] The computer-readable recording medium according to the present invention is characterized in that a bit array that represents a combination of a plurality of mixing ratios of color materials in a binary number is a gene, and is a function of at least one characteristic quantity representing a color difference. Set an evaluation index,
For a gene population consisting of a plurality of genes, evaluating the genes in the gene population using the above-mentioned evaluation index, storing the best-evaluated gene, and determining that the gene with the optimal combination has not been obtained. Next, in the above-mentioned gene population, the higher the evaluation of the gene, the higher the probability, the step of selecting the gene to be left, the step of crossing over the left gene, and the step of further mutating the gene. And returning to the step of performing the above-described evaluation on the gene population thus obtained, and recording a color matching program that repeats the above-described processing until an optimal combination is obtained.

[0008]

Embodiments of the present invention will be described below with reference to the accompanying drawings. In computer color matching, a difference between a target color and a color of a color material mixture calculated by a computer is evaluated. Here, calculation for evaluating the color difference is performed so that the evaluation function G is minimized, and the combination of the color materials and the mixing ratio are determined. G = √ in ^{((α · IS) 2 +} (β · M) 2 + (γ · D) 2) (1) This equation (1), IS is a function of the spectral reflectance difference △ R _lambda, values and that the sum of the square value of the reflectance difference of each wavelength, and the like following values summed by weighting at reflectance difference △ R _lambda to the weighting function f _lambda of each wavelength as shown in equation. IS = (Σ ( _fλ · △ _Rλ ) ² ) M is a function of metamerism, and is a conditional color matching index M under a light source different from the designated light source. If the different light sources have more than one obtained by the following equation with respect to conditions such as Color Index M _i of each light source. M = √ (ΣM _i ² ) D is a function of a color value, and is a value obtained by weighting each component with respect to the brightness difference ΔL, the saturation difference ΔC, and the hue difference ΔH, and combining them as a color difference. It is. D = √ ((f _L · △ L) ² + (f _C · △ C) ² + (f _H · △ H) ² ) where f _L , f _C , and f _H are weighting functions of each component. is there. (In the calculation example described later, f _L = f _C = f _H =
1.0 is set. ) The weight of each component can be freely changed. The evaluation index G is calculated based on these spectral reflectance differences I.
S, metamerism function M, contribution rate α, β of color function D,
is determined so that γ can be changed freely.

In the case of metallic color, pearl color, etc., the color looks different depending on the angle. In such a case, when the multi-angle spectrophotometer is used at a plurality of angles, the evaluation index G between the target color and the color of the colorant compound calculated by the computer is calculated as the evaluation index G for each angle. seeking, and in consideration of the contribution rate f _a of the angle _{G = √ (Σ (f a} · G a) 2). In addition, in the expression (1) of the evaluation index G, the flop value F and the contribution ratio δ can be added as the characteristic values representing the fourth color difference.

In the present embodiment, a genetic algorithm is applied as a technique for obtaining an optimum colorant composition. In the genetic algorithm, the mixture ratio of color materials is expressed as a gene by applying a gray code and expressed as a gene.
Through the process of gene crossover / gene mutation, the best gene is selected using the evaluation index G.
The above evaluation index G is applied to the evaluation of this gene. This method makes it possible to find an optimal solution that was difficult to find in the past.

Next, the color matching processing by the genetic algorithm will be described. The term "gene" used here means
A variable is a binary number. In the case of computer color matching, the combination of color materials (primary colors) is treated as a binary number (gray code). The content of the gene is a bit array of 0 or 1, and the larger the number of bits, the more significant digits of the variable. For example, when a mixture is geneticized with 14 bits per primary color, 100/2 ¹⁴ = 0.006 in the case of% combination
The calculation result can be obtained with the precision of For the white primary color, white primary color% = 100−Σchromatic color%, so only the chromatic color combination of the color material used is set in the gene. For example, when the number of chromatic colors is 4, the bit length of the gene is 14 bits * 4 primary colors = 56 bits. In the following example gene sequence, the first 14 bits represent the first primary color, the next 14 bits represent the second primary color, the next 14 bits represent the third primary color, and the last 14 bits represent the first primary color. 4 represents the primary colors. In the genetic algorithm, there are two methods, a method of treating a binary code as a gene as it is and a method of treating a binary code as a gray code. Here, it is assumed that a gray-coded gene is handled.

FIG. 1 shows a configuration of a color matching system using a genetic algorithm. This system is C
A computer 10 including a PU, a RAM, a ROM, and the like;
An input device such as a display device 12, a keyboard 14, a mouse 16, an output device such as a printer 18,
A device that handles recording media such as a hard disk device 20 and a CD-ROM device 22 is provided. This configuration in which the colorimeter 24 for inputting color data is connected is the same as that of a normal computer system. The computer 10 performs a color matching process according to a user's instruction from the input device, and the result is displayed on the display device 12.

FIG. 2 is a flowchart of a computer color matching process using a genetic algorithm. This flow is executed by the computer 10 in FIG. 1, and the computer program is recorded on a recording medium such as the hard disk device 20 and the CD-ROM device 22. Hereinafter, the contents of the computer color matching program will be described. Note that various initial values in the following processing are set using a default file.

First, an initial population of genes is created (S1).
0). The number of genes (for example, 100) in the population is predetermined. The content of the initial gene is determined by random numbers,
In addition, a 100% combination of each primary color of the coloring material and a combination determined by an initial value calculation using the least squares method for the spectral reflectance difference IS are put in an initial population of genes. For the gene, a chromatic color mixture% is set in gray code. The white primary color is calculated as the white primary color% = 100−Σchromatic color%, and the white primary color portion is not written in the gene. Table 1 shows an example of the initial gene population when five chromatic coloring materials are used. Here, the initial value of 2 corresponds to the color material of the white primary color.

[Table 1]

Next, it is determined whether or not a predetermined number of loops has been exceeded (S12). The predetermined number of loops is, for example, 1000. If the number of loops exceeds the predetermined number, this calculation ends. If the number does not exceed the predetermined number of loops, the gene is evaluated using the evaluation index G (S14). here,
The smaller the evaluation index G, the higher the evaluation. So, ex
The evaluation is performed using p (−G). This evaluation is performed for all genes in the population. However, if the sum of chromatic% exceeds 100%, the evaluation is set to the lowest value. Then, the gene with the highest evaluation is stored (S16).

Next, convergence determination is performed on the data of the gene with the highest evaluation (S18). If the replacement of the gene with the highest evaluation has not been performed a predetermined number of times or more for the convergence determination even after repeating the loop, the calculation is terminated assuming that the optimal combination has been obtained (converged).

If it is determined that the gene does not converge, a table for selecting the gene to be left next is created (S2).
0). Here, assuming that the number of genes in a group is N and the evaluation of gene i in the group is f _i , the cumulative probability q _{i of} gene i is
= Is calculated as follows. Sum of evaluations F = Σf _i (i = 1, N) Cumulative probability of gene i q _i = Σf _j / F (j = 1, i) This is repeated from i = 1 to N. q _i is a real number from 0 to 1.

Next, a gene to be left is selected using the probability of the gene (S22). The value of random number r (random number from 0 to 1) is q
If you are entering a port to _{i-1 ~q i,} selects a gene i. Since the range of the higher rated gene q _i-1 ~q _i increases,
The probability of being selected increases. This operation is repeated N times.
As a result of this operation, the number of genes in the population does not change,
The likelihood of duplicated genes with high evaluation increases, and the evaluation of the entire population improves.

Next, gene crossover is performed (S24). The crossover probability is predetermined. If random number r (random number from 0 to 1) ≦ crossover probability, gene i is selected as the parent. this,
Repeat for i = 1, N. If the number of selected parent genes is odd, the last selected gene is discarded, and the number of parents is made even. Then, two parent genes are paired, and bit sequences are exchanged at random positions by random numbers. To explain one example, in the next two parent genes, the exchange position is determined by random numbers, and the sequence behind the exchange position is exchanged. Parent gene 1: 001001110101000110 Parent gene 2: 0001011011101101110 In this example, the exchange position is the eighth bit from the top. As a result, the new gene is as follows. New gene 1: 001001111101101110 New gene 2: 00010101011000110

Next, the gene is mutated (S2).
4). Here, the mutation probability is determined in advance. If the random number r (random number of 0 to 1) ≦ mutation probability is satisfied for the bit sequence of the gene, the bit is inverted.
This operation is performed for all bits of the gene. This is done similarly for all genes in the population. Inclusion of the gene due to the mutation increases the likelihood of obtaining a true solution rather than a metastable solution. Finally, return the best gene mix. Then, the process returns to step S12.
To explain one example of mutation, for the next parent gene,
For example, for the eighth bit from the beginning, the bits are inverted to obtain a new gene 2. Parent gene 1: 00100111101101110 New gene 2: 00101100101000110

Next, sample 1 (dark brown) and sample 2
(Red) as a target color, the designated light source for color matching is a D-65 light source, and a 10 ° field of view is used.
Computer color matching calculations were performed. In sample 1, white, red-2, red-3, red-4, brown-1, blue-
1. Seven kinds of color materials of Pearl-1 are used. In sample 2, six kinds of color materials of white, red-1, red-2, blue-1, yellow-1, and pearl-2 are used. In these samples, a plurality of similar color materials are used. Here, the following evaluation index GG = √ ((α · IS) ² + (β · M) ² + (γ · D) ² ) α = 0.1, β = 0.0, γ = 0.9 Was used to perform color color matching calculations. Here, the color value function D is the most important. In addition, metamerism color matching calculation (α = γ = 0.
0), isometric color matching calculation (β = γ =
0.0). Tables 1 and 2 show the results. 3 to 5 are graphs of the reflectance curvature and the metamerism color matching calculation result, the graph of the reflectance curvature and the isometric color matching calculation result, and the color matching calculation result using the reflectance curvature and the evaluation index G for the sample 1. It is a graph of. Similarly, FIGS. 6 to 8 show graphs of the reflectance curvature and the metamerism color matching calculation result for Sample 2, graphs of the reflectance curvature and the isometric color matching calculation result, and the color matching calculation using the reflectance curvature and the evaluation index G. It is a graph of a result. In these figures,
The broken line shows the reflectance curve, and the solid line shows the calculation result.

[0022]

[Table 2]

[0023]

[Table 3]

According to the calculation result, in the metamerism color matching calculation (α = γ = 0.0), the color difference under the designated light source is small, but the color difference with other light sources is large, so that the metamerism occurs. are doing. In the isometric color matching calculation (β = γ = 0.0), the reflectance difference is small, but the color difference remains. By performing the color matching calculation using the evaluation index G (α = 0.1, β = 0.0, γ = 0.9), the best result is obtained, the color difference under the designated light source is small, In addition, it is possible to obtain a color material mixing ratio with less metamerism.

Note that the conventional computer color matching is usually used to determine the combination of three color materials, and is used for both the metamerism color matching calculation and the isometric color matching calculation.
When four color materials were used, a highly accurate result was not obtained, and reproducibility was poor. Therefore, the results of blending a large number of coloring materials as in Samples 1 and 2 could not be obtained. In the present invention, both metamerism color matching calculation and isometric color matching calculation are possible. further,
The blending result can be calculated using the general evaluation index G.

In the present embodiment, by freely setting the evaluation index G, it is possible to perform computer color matching that matches the senses and desires of the toner even for colors that have been difficult to achieve. For example, when there is no color material used for the target color among the color materials on hand, or when the condition-equivalent color cannot be obtained or when the color match cannot be performed, it is necessary to determine the combination and the mixing ratio of the color material that becomes the closest color. In such a case, the conventional color difference Δ
A method that considers only E = √ (△ L ² + △ C ² + △ H ² ) often produces a calculation result that differs from the sense of a toned person. However, in the present embodiment, in such a case, by adjusting the color function D and the metamerism function M of the evaluation index G, it is possible to obtain a combination and a mixing ratio of a color material desired by a toner. Also, when using a multi-angle spectrophotometer, by increasing the contribution of G _a of the desired angle, it has become possible to determine the combination and mixing ratio of the coloring material that suits by the desired angle.

The method of using the least squares method for the reflectance difference and the method of setting the difference of the color value to 0 are effective in the conventional isomeric matching and metamerism matching, but the evaluation index G is minimized. It is not suitable for finding the mixing ratio of the coloring material to be used. In addition, a blending search method using the dichotomy or the shotgun method has been attempted, but a metastable solution may be obtained, and the blending ratio of the coloring material having the minimum evaluation index is not always obtained. In the present embodiment, an optimal mixture ratio is obtained by using a genetic algorithm.

[0028]

According to the present invention, a true solution can be obtained instead of a metastable solution by using a genetic algorithm in computer color matching. By freely setting the evaluation index, computer color matching that matches the senses and wishes of the toned person can be performed even for colors that have been weak until now, using an evaluation method that matches the feeling of the toned person. It becomes possible. As a result, a more practical combination of color materials and a compounding ratio thereof are required.

[Brief description of the drawings]

FIG. 1 is a diagram of a computer system.

FIG. 2 is a flowchart of a computer color matching process using a genetic algorithm.

FIG. 3 is a graph of a reflectance curvature and a metamerism matching calculation result for a sample 1;

FIG. 4 is a graph of a reflectance curvature and an isomeric matching calculation result for a sample 1;

FIG. 5 is a graph of a matching calculation result using a reflectance curvature and an evaluation index for Sample 1;

FIG. 6 is a graph of a reflectance curvature and a metamerism matching calculation result for a sample 2;

FIG. 7 is a graph of a reflectance curvature and an isometric matching calculation result for Sample 2;

FIG. 8 is a graph of a matching calculation result using a reflectance curvature and an evaluation index for Sample 2;

[Explanation of symbols]

10 computer, 12 display device,
20 hard disk drive, 22 CD-ROM
apparatus.

──────────────────────────────────────────────────続 き Continued on the front page (51) Int.Cl. ⁷ Identification symbol FI Theme coat ゛ (Reference) C12N 15/09 C12N 15/00 A F-term (Reference) 4B024 AA11 CA01 CA11 HA19 4B029 AA07 AA23 BB20 FA10 5B057 AA10 CA01 CA08 CA16 CB01 CB08 CB16 CE11 CE17 DB02 DB06 DC25 5C077 MP08 PP31 PP33 PP43 PP47 PP57 PQ12 5C079 KA15 LA02 LA31 MA01 MA11 NA15

Claims (7)

[Claims] 1. A combination of a plurality of coloring materials in a combination ratio of
A bit array represented by a base is expressed as a gene, and an evaluation index that is a function of at least one characteristic amount representing a color difference is set. For a gene group including a plurality of genes, the genes in the gene group are evaluated by the evaluation index. Is evaluated using, and the step of storing the best-evaluated genes, If it is determined that the optimal combination of genes has not been obtained, then, in the gene population, the higher the evaluation of the gene, the higher the probability And selecting a gene to be left, then crossing over the left gene, and further mutating the gene, and performing the above-described evaluation on the gene population thus obtained. And a computer color matching method for repeating the above processing until an optimum combination is obtained. 2. The computer color according to claim 1, wherein the function of the characteristic amount representing the color difference includes at least one of a spectral reflectance difference function, a metamerism function, and a color function. Matching method. 3. The evaluation index is at least two of a function of a spectral reflectance difference, a metamerism function, and a color function.
2. The computer color matching method according to claim 1, wherein the two are weighted and added. 4. The function according to claim 2, wherein the function of the characteristic quantity representing the color difference further includes a flop value.
Computer color matching method described in 1. 5. The computer color matching method according to claim 1, wherein the evaluation index is determined in consideration of a contribution ratio of each angle based on a value obtained for each of a plurality of angles. Method. 6. The computer color matching method according to claim 1, wherein the initial gene population includes a 100% composition of each colorant and a composition obtained by an initial value calculation based on spectral reflectance data. 7. A combination of a plurality of coloring materials in a combination ratio of 2
A bit array represented by a base is expressed as a gene, and an evaluation index that is a function of at least one characteristic amount representing a color difference is set. For a gene group including a plurality of genes, the genes in the gene group are evaluated by the evaluation index. Is evaluated using, and the step of storing the best-evaluated genes, If it is determined that the optimal combination of genes has not been obtained, then, in the gene population, the higher the evaluation of the gene, the higher the probability And selecting a gene to be left, then crossing over the left gene, and further mutating the gene, and performing the above-described evaluation on the gene population thus obtained. And a computer-readable recording medium for recording a color matching program for repeating the above processing until an optimum combination is obtained.