High-precision color separation model of color adaptive coding multicolor printer

Abstract

According to the high-precision color separation model of the color adaptive coding multi-color printer, the color separation model of each partition is established in each sub-color gamut by adopting a multi-color term fitting method, and on the basis, middle mixed color local replacement, mixed color replacement and light ink replacement are improved to optimize the models. The method comprises the following steps of: 1, establishing a sampling sample in each sub-color gamut of a multi-color printer, and constructing a three-color color separation model of each subarea by adopting a multi-color item fitting method; secondly, establishing a middle mixed color local replacement relationship and a mixed color replacement relationship based on an establishment method of a grey balance equation, further performing color separation on a color separation result of each partition by adopting a middle mixed color local replacement equation and a mixed color replacement equation, and optimizing the model; thirdly, a critical domain relation between shallow ink and deep ink is established, and when the usage amount of the deep ink is lower than a critical value, the shallow ink is used for replacement printing; and when the use amount of the deep ink is higher than the critical domain, the deep ink is directly used, so that a better and finer color separation effect is achieved.

Description

Technical Field

[0001] The present application relates to a high-precision color separation model for a multi-color printer, and in particular to a high-precision color separation model for a color adaptive coding multi-color printer, and belongs to the technical field of printing and packaging. Background Art

[0002] A multi-color printer is a printer that adds other ink channels (such as orange, green, and blue ink channels) to the four-color channels of a CMYK printer. There are two types of printers: one that can increase the color output range of the printer, thereby reproducing brighter colors; and the other that can use light ink to improve the printing quality of tones, making the printing more delicate, richer in layers, and smoother in tones. While multi-primary color printers improve the printing quality, they also cause defects such as ink spots, which increases the gap between the input and output of the printer by 3. For multi-color printers, their color profiles describe the color reproduction capabilities of the printer, and are the basis for color conversion and color control to achieve high-quality printing. The core of the printer color profile is the color model for mutual conversion between the device-independent color space and the printer output color space. However, there are not many research and developments on multi-color separation models at present, and the existing multi-color separation schemes involve technical confidentiality and patents, and are not disclosed. The research and development of multi-primary color printer separation models focuses on the above two types of printers, while there is less research and development on printer separations that combine the characteristics of these two printers and have both expanded color gamut and improved tonal printing quality. Therefore, it is urgent to combine the research and development of the two types of printers, rely on the ten-color inkjet printer, develop a color separation model for a printer that combines the advantages of these two types of printers, and optimize the color separation technology involved in the color control system of this printer.

[0003] The research and development of color separation models in existing technologies can be divided into two categories: color separation research and development based on color chromaticity and color separation research and development based on color spectral distribution data. The color separation research and development of multi-color printers based on chromaticity is the transformation from device-independent color space to the dot area rate of each color channel of the printer, which is the modeling process from three-dimensional data to multi-dimensional data; while the research and development of color separation of printers based on spectrum is the transformation from the spectral data of the color to the ink volume of each color channel of the printer, which is the modeling process from multi-dimensional data (narrow-band spectral data within the 31-dimensional visible spectrum range) to multi-dimensional data. This technology has very high requirements for the accuracy of data sampling and processing, and also has extremely high technical requirements for operators. If the spectral data cannot meet the accuracy requirements, the effect of color reproduction will be greatly reduced. Therefore, spectral color reproduction technology has not yet been widely used in practice.

[0004] In the color separation technology of multi-primary color printers based on color chromaticity, the model method uses the Niegber model to simulate the output process of the printer. Only a small number of samples are needed to build the model. However, since the color reproduction process of the printer is affected by factors such as the substrate, ink, and equipment accuracy, it is difficult for the Niegber model to accurately predict the printer output. The artificial neural network method improves the accuracy of the model through a large number of sample training, which is time-consuming and not very practical in practical applications. The multidimensional lookup table method has a complex table building process, and the accuracy of the model depends on the sampling density and interpolation operation accuracy.

[0005] For printers with light ink channels, since the chroma of light ink and dark ink is basically the same, but the saturation difference is significant, it is impossible to use color adaptive coding theory for color separation. At present, according to the color partition theory, multivariate color gamut regression analysis is used in each partition to establish a color separation model, and the color mixing replacement method, medium color mixing replacement and light ink mapping method are combined to obtain the color separation scheme of the hybrid printer. There are few comprehensive research and developments. This application conducts systematic research and development on this.

[0006] The problems that need to be solved in the color separation of multi-color printers in the prior art and the key technical difficulties of this application include:

[0007] (1) The existing technology of spectral-based printer color separation is the conversion from the spectral data of the color to the ink volume of each color channel of the printer, which is a modeling process from multidimensional data to multidimensional data. However, this technology has very high requirements for the accuracy of data sampling and processing, and also has extremely high technical requirements for operators. If the spectral data cannot meet the accuracy requirements, the effect of color reproduction will be greatly reduced. Therefore, spectral color reproduction technology has not yet been widely used in practice. At present, there is not much research and development on multi-color color separation models, and the existing multi-color color separation schemes involve technical confidentiality and patents, and are not disclosed. The research and development of multi-primary color printer color separation models focuses on two types of printers, while there is less research and development on printer color separation that combines the characteristics of these two types of printers and has both expanded color gamut and improved color tone printing quality. Therefore, it is urgent to combine the research and development of the two types of printers, rely on ten-color inkjet printers, and develop color separation models for printers that combine the advantages of these two types of printers, and optimize the color separation technology involved in the color control system of such printers.

[0008] (2) In the color separation technology of multi-primary color printers based on color chromaticity, the model method uses the Niegber model to simulate the output process of the printer. Only a small number of samples are needed to build the model. However, since the color reproduction process of the printer is affected by factors such as the substrate, ink, and equipment accuracy, the Niegber model is difficult to accurately predict the printer output. The artificial neural network method uses a large number of sample training to improve the model accuracy, which is time-consuming and not very practical in practical applications. The multidimensional lookup table method has a complex table building process, and the accuracy of the model depends on the sampling density and interpolation operation accuracy. For printers with light ink channels, since the chromaticity of light ink and dark ink is basically the same, but the saturation difference is significant, it is impossible to use the color adaptive coding theory for color separation. At present, according to the color partitioning theory, multivariate color gamut regression analysis is used in each partition to establish a color separation model, and the color mixing replacement method, medium color mixing replacement and light ink mapping method are combined to obtain the color separation scheme of the hybrid printer. There is little comprehensive research and development on this, and it is urgently needed to carry out systematic research and development.

[0009] (3) The prior art lacks a method for establishing a color separation model for each partition based on color coding using a multi-color item fitting method in each sub-color domain of a multi-color printer, and lacks improved local displacement of mixed colors, mixed color displacement, and light ink displacement to optimize the model. First, in each sub-color domain of a multi-color printer, no sampling samples are established, and there is a lack of a method for multi-color item fitting to construct a three-color color separation model for each partition; second, there is a lack of an establishment method based on a gray balance equation, and the local displacement relationship of mixed colors and the mixed color displacement relationship are not established, so it is impossible to use the local displacement equation of mixed colors and the mixed color displacement equation to further separate the color results of each partition and optimize the model; third, the critical domain relationship between light ink and dark ink is not established, and when the amount of dark ink used is lower than the critical value, light ink displacement printing cannot be used; this results in low color separation accuracy for multi-color printers, which is not conducive to the development of color separation for multi-color printers. Summary of the invention

[0010] This application is the research and development of printer color separation that has both expanded color gamut and improved color tone printing quality. Based on a ten-color inkjet printer, a color separation model is developed for a printer that combines the advantages of the two types of printers, and the color separation technology involved in the printer color control system is optimized. Based on color adaptive coding, the color separation model of a multi-color inkjet printer is self-encoded for the color gamut composed of the six dark inks C, M, Y, K, O, and G of the ten-color printer, and a color separation model is established to obtain the color separation values ​​of each partition. The mapping relationship between dark ink and light ink is used to complete the conversion of the 6-color color separation value to the ten-color color separation value, and the entire color separation modeling of the ten-color channel of the printer is completed. A multi-color item fitting algorithm is used to construct a color separation model in each sub-partition; referring to the mid-mixed color replacement process (non-color structure process), the mid-mixed color local replacement and mixed color replacement are developed, and their role in the multi-color printer color separation model is determined; the light ink replacement method is developed to determine its influence on the multi-color printer color separation model, and a high-precision and high-quality color separation model of a multi-color printer with color adaptive coding is established.

[0011] In order to achieve the above technical effects, the technical solutions adopted in this application are as follows:

[0012] Color adaptive coding multi-color printer high-precision color separation model. For multi-color printers with added light ink, a replacement critical domain between light ink and dark ink is established on the basis of the four-color separation mechanism to achieve color separation on the entire printer channel. The printer color gamut is self-encoded, and empirical models are constructed in each sub-gamut to establish the color separation model of each partition. The printer color separation is completed by using the local replacement of mixed colors and the replacement of mixed colors.

[0013] Color separation model for multi-color inkjet printers based on color adaptive coding: self-encode the color gamut composed of the six dark inks C, M, Y, K, O, and G of a ten-color printer, establish a color separation model to obtain the color separation values ​​of each separation area, and use the mapping relationship between dark ink and light ink to complete the conversion of the six-color separation value to the ten-color separation value, thus completing the entire color separation modeling of the ten-color channel of the printer;

[0014] The process of establishing the color separation model of a multi-color printer based on color adaptive coding is as follows: after verifying that the chromaticity of light ink is consistent with that of dark ink, the light ink channel of the ten-color printer is separated, and the color gamut of the printer composed of six dark inks is color adaptively encoded according to color adaptive coding, and self-encoded into five subspaces. In each subspace, a subspace color separation model is established using a multi-color term regression equation, a gray balance equation is established, and a local replacement rate of medium mixed colors is set to obtain a color separation result of local replacement of medium mixed colors. The mixed color part of the original is mixed by two primary colors in a certain proportion. A mixed color replacement equation is established based on gray balance. The obtained mixed color replacement equation is the result of complete replacement of the mixed colors. The replacement rate is set according to the original to obtain a color separation result of mixed color replacement. For the color part with less dark ink, the corresponding light ink is used to obtain a better color reproduction effect, and a value relationship between light ink and dark ink is established, that is, when the use of dark ink is lower than the critical domain, light ink is used to replace it, and when it is higher than the critical domain, dark ink is used, so that the color separation range is expanded to the entire ten-color channel of the printer to obtain a better color separation effect.

[0015] Preferably, based on color adaptive coding, a multi-color item fitting method is used in each sub-color gamut of a multi-color printer to establish a color separation model for each partition, and on this basis, the medium mixed color local displacement, mixed color displacement and light ink displacement are improved to optimize the model: first, in each sub-color gamut of the multi-color printer, a sampling sample is established, and a three-color color separation model for each partition is constructed by a multi-color item fitting method; second, based on the establishment method of the gray balance equation, a medium mixed color local displacement relationship and a mixed color displacement relationship are established, and the medium mixed color local displacement equation and the mixed color displacement equation are used to further separate the color of the color separation results of each partition to optimize the model; third, a critical domain relationship between light ink and dark ink is established, and when the dark ink usage is lower than the critical value, light ink is used for displacement printing; when the dark ink usage is higher than the critical domain, dark ink is used directly.

[0016] Preferably, the sample color adaptive coding is prepared by calculating the hue angles of the six dark inks of the printer, adaptively encoding the printer colors into five areas of YOK, OMK, CMK, GCK, and GYK, sampling is performed in each color separation area to form a sample set for modeling each partition, in the GCK partition, the three colors G, C, and K are overprinted in the range of 0-100% dot area rate with a step size of 10% to form a sample set for each partition, the printer color management is turned off, and the sample colors of the five partitions are directly printed out through the printer channel control.

[0017] Preferably, multivariate color gamut regression analysis: the conversion model of the color separation model from the CIELAB color space to the CMYK color space belongs to multivariate multiple nonlinear regression, and the relationship between the original color space and the target color space is established through multi-color terms, and the coefficients are obtained by the least squares method to determine the mathematical conversion model of the two color spaces, and finally the target color space and mixed color are calculated by the multi-color term relationship, and the decanate multi-color terms are used to establish the color separation model for each separation.

[0018] Preferably, the conversion model from CIELAB color space to CMYK color space is a multivariate multiple nonlinear regression problem. Based on the correlation between multiple dependent variables and multiple independent variables, the following mathematical model is established for the multiple univariate linear regression problem: Assume that there are n independent variables x 1至n , corresponding to m dependent variables y 1至m , establish the relationship between the two sets of data:

[0019]

[0020] where β ij (i=0-n;j=0-m) are unknown parameters, ε j (j=0-m) is a random error, which is not independent of each other. Assuming that these error values ​​obey a multivariate normal distribution, the above model can be written in the form of a matrix as follows:

[0021]

[0022] The least square method is used to solve the coefficients of the multiple univariate regression equation and establish the regression relationship:

[0023] (x 11 , x 12 , …, x 1n ,;y 11 ,y 12 , …, y 1f )

[0024] (x 21 , x 22 , …, x 2n ,;y 21 ,y 22 , …, y 2f )

[0025] ……Formula 3

[0026] The multivariate multiple linear regression model is expressed as:

[0027]

[0028] ε is the error matrix. The error value is neglected and the coefficient matrix is ​​solved by the least square method to obtain the multivariate multiple linear regression equation. The color separation model of the printer is converted from the CIELab color space to the CMY color space. The dependent variable is the dot area rate of the three colors of yellow, magenta and cyan. The independent variables are L, a, b and the product terms formed by each other. Multivariate multiple nonlinear regression analysis is used to obtain the values ​​of other independent variables when the values ​​of the independent variables L, a, and b are known. The nonlinear regression problem is converted into a linear regression problem to simplify the calculation process.

[0029] Preferably, the model is established by printing out the sample sets of each partition, calculating the Lab value of the sample color as the basic data for establishing the model, taking the Lab value as the independent variable and the dot area rate of the primary color of each partition as the dependent variable to establish a multi-color regression equation, and based on the multivariate color gamut regression analysis with GCK partition, the established multi-color regression equation is as shown in Formula 5:

[0030]

[0031] G, C, K represent the dot area ratio of the three primary colors in the partition, σ ij (i=1-3; j=0-9) represents the multi-color coefficient to be calculated. For all sample colors in the partition, equations are established. Let P 1 =L,P 2 =A,P 3 =B,P 4 =L 2 , P 5 =A 2 , P 6 =B 2 , P 7 =LA,P 8 =AB,P 9 =LB, transform the multivariate nonlinear regression problem into a multivariate linear regression problem, and obtain the matrix equation based on all sample colors as shown in Formula 6:

[0032]

[0033] Calculate σ by multiple linear regression analysis ij The least squares estimate of the partition is finally obtained, and the mathematical conversion model from the chromaticity value of the partition to the dot area ratio value is obtained. The multi-color regression color separation model is established for YOK, OMK, CMK, and GYK partitions in the same way. The regression model established for KCG partition is as shown in Formula 7:

[0034]

[0035] Preferably, the medium mixed color is partially replaced and then the color is separated: the medium mixed color is partially replaced according to the original manuscript requirements, first the gray balance curve is established for the CMYK4 colors, and then the medium mixed color replacement rate μ kPerform local replacement;

[0036] Calculate the density of each color block under the R filter, G filter, and B filter respectively, calculate multiple times to get the average value, first completely replace the mixed color in the middle to determine the relationship between K and CMY;

[0037] D YB , D MB , D CB They are the densities of the three color inks Y, M, and C measured under the blue filter, D YG , D MG , D CG They are the density of Y, M, and C inks measured under the green filter, D YR , D MR , D CR are the densities of the three color inks Y, M, and C measured under the red filter, D k It is the visual density of monochrome black ink. It is calculated according to the equivalent neutral density calculation method to obtain the proportional coefficient of yellow, magenta and cyan inks when the neutral gray balance is achieved at each level; D e To achieve gray balance, the gray density value under each color filter is used to calculate the proportional coefficient of each level of the ladder scale, and the proportional coefficient of each level is multiplied by the corresponding main density to obtain the density values ​​of cyan, magenta, and yellow required to form each level of neutral gray balance; the dot area rate of each color block is calculated, and the empirical correction index is taken as 2;

[0038] When the number of multi-color items is 3, a good fit is achieved, and the equivalent substitution relationship between K and CMY is obtained:

[0039]

[0040] The displacement relationship is the result of replacing all mixed colors with black ink. The local displacement rate of mixed colors in the local displacement setting is μ k , the color separation value after all the mixed colors are replaced is C O , G O , K O , the K value of the KCG partition to be replaced is K O *μ k , and get the equivalent C K , M K ,Y K Color separation value, the number of color separations becomes CMYGK5 colors, the color separation value is C e 、M e , Y e , G e , K e , after the black ink in the GCK partition is partially replaced:

[0041]

[0042] The final color separation result is shown in Formula 9.

[0043] Preferably, the color mixing replacement and color separation: orange O and green G are medium mixed colors of the ten-color printer. Based on the gray balance curve establishment method, the orange O is overprinted with yellow Y and magenta M in appropriate proportions. If the color can be reproduced the same as the single-color orange O, the density of the single-color orange O is calculated at this time. The density is the density of the orange O equivalent to MY at the dot area rate at this time. The density is converted into the dot area rate to establish the color replacement relationship between O and MY: first, the single-color gradient measurement and control strips of O, M, and Y are designed, with a step unit of 5%, and a total of 20 color blocks with dot area rates ranging from 0% to 100%. The density of each sampling point in the measurement and control strip under the G filter and the B filter is measured;

[0044] According to the calculated density values ​​of each sampling point, the following equations are obtained between O and MY:

[0045]

[0046] Where D OG , D OB are the main density values ​​of color O measured under the green and blue filters, ω Y ,ω m D is the ratio coefficient of M and Y when forming the corresponding tone density of O. YG , D MG Refers to the density value of M and Y under the green filter; DYB and DMB refer to the density value of M and Y under the blue filter. The solution of the proportional coefficient is shown in formula 11:

[0047]

[0048] Obtain the M and Y ratio coefficients required to form each level of O color, and multiply them with the main density to obtain the actual required M and Y density as shown in formula 12:

[0049]

[0050] D Ye , D Me The actual effective density values ​​of Y and M required to form the color mixing O, F Y , F M In order to obtain the effective dot area rate based on the effective density value using the Yule-Nielsen formula, a univariate multivariate color gamut regression analysis is used to determine the coefficients of the multicolor items and obtain the equivalent substitution relationship between O and MY as shown in Formula 13:

[0051]

[0052] The equivalent substitution relationship between G and CY is obtained as follows:

[0053]

[0054] Formula 13 and Formula 14 are equivalent color replacement values ​​when the 0 color and the G color are replaced to the maximum extent. The mixed color replacement rate is determined according to the original manuscript. The replacement rate when the O color is partially replaced by the M and Y colors is α, and the replacement rate when the G color is replaced by the C and Y colors is β, as shown in Formula 15 and Formula 16:

[0055]

[0056] Similarly, based on the KCG partitioning and the local replacement of the mixed color, the local replacement of the mixed color is performed again, and the final color separation result is formula 17:

[0057]

[0058] The local displacement rate of mixed color in the local displacement setting is μ k , the color separation value after all the mixed colors are replaced is C O , G O , K O , the K value of the KCG partition to be replaced is K O *μ k , and get the equivalent C K , M K ,Y K Color separation value, the number of color separations becomes CMYGK5 colors, the color separation value is C e 、M e , Y e , G e , K e .

[0059] Preferably, the light ink replacement method is used for color separation: the light ink and the dark ink have the same hue, and a linear relationship is established between the two. C ,I M ,I K ,I LK ,I LLK ,I LM ,I LC The dot area ratios of magenta, cyan, black, light black, light light black, light magenta, and light cyan are expressed by taking the density of the dark ink closest to the density when the light ink concentration is 100% as the basic data to establish the substitution relationship between the light ink and the dark ink;

[0060] According to the replacement relationship between light ink and dark ink, I K =aI LK ,I K =bI LLK ,I C =eI LC ,I M =fI LMThe linear mapping relationship of , formula 18 is the replacement relationship between black ink and light black ink, light light black ink;

[0061]

[0062] a, b are the color separation critical domain of K. Print out the black and light ink gradation scale, calculate the density value of each color block, the density when the light black ink concentration is 100% is close to the density when the dark black ink concentration is 20%, let the critical domain of the two channels be 0.2, the value of b is 0.2, and the value of a is 0.45. When the black color separation value is less than the critical value of 0.2, let the black printing channel color separation value be 0, replace the black channel with light black, and calculate its color separation value k / 0.2 according to the linear transformation; if the black color separation value is greater than or equal to 0.2 and less than 0.45, let the black printing channel color separation value be 0, replace the black channel with light black, and calculate its color separation value I according to the linear transformation K / 0.45;

[0063] The substitution relationship between cyan and light cyan, and between magenta and light magenta are shown in formulas 4-18 and 4-19. The value of e is 0.3, and the color separation value of f is 0.25. When the cyan color separation value is greater than 0.3, the color separation value of the cyan channel is set to 0, and replaced by the color separation value of the light cyan channel Ic / 0.30; when the magenta color separation value is greater than 0.25, the color separation value of the magenta channel is set to 0, and replaced by the color separation value of the light magenta channel IM / 0.25.

[0064]

[0065] According to the results of medium mixed color local replacement and mixed color local replacement, the light ink replacement is performed based on the critical domain of light ink replacement, and finally the color separation data within the ten color channels of the printer is obtained.

[0066] Compared with the prior art, the innovations and advantages of this application are:

[0067] (1) Based on color adaptive coding, this application uses a multi-color item fitting method in each sub-color gamut of a multi-color printer to establish a color separation model for each partition, and on this basis, improves the local displacement of medium mixed colors, mixed color displacement, and light ink displacement to optimize the model. First, in each sub-color gamut of a multi-color printer, a sampling sample is established, and a three-color color separation model of each partition is constructed by a multi-color item fitting method; second, based on the establishment method of the gray balance equation, a local displacement relationship of medium mixed colors and a mixed color displacement relationship are established, and the color separation results of each partition are further separated by using the local displacement equation of medium mixed colors and the mixed color displacement equation to optimize the model; third, a critical domain relationship between light ink and dark ink is established. When the amount of dark ink used is lower than the critical value, light ink is used for displacement printing; when the amount of dark ink used is higher than the critical domain, dark ink is used directly. This application systematically integrates the local displacement of medium mixed colors, mixed color displacement, and light ink displacement into the color separation model of a multi-color printer, which is conducive to improving the color separation accuracy of a multi-color printer. It has a guiding and reference role in the future research and development of color separation of multi-color printers.

[0068] (2) For a multi-color printer with added light ink, the present application establishes a replacement critical domain between light ink and dark ink on the basis of a four-color color separation mechanism, realizes color separation on the entire printer channel, performs self-encoding on the printer color gamut, constructs empirical models in each sub-color gamut to establish a color separation model for each partition, and completes the printer color separation by using medium mixed color local replacement and mixed color replacement; after verifying that the chromaticity of the light ink is consistent with the chromaticity of the dark ink, the light ink channel of the ten-color printer is separated, and the printer color gamut composed of the six-color dark ink is color-adaptively encoded according to color adaptive coding, and self-encoded into five subspaces. In each subspace, a subspace is established by using a multi-color term regression equation. The intermediate color separation model is established, the gray balance equation is established, the local replacement rate of the medium mixed color is set, and the color separation result of the local replacement of the medium mixed color is obtained. The mixed color part of the original is mixed by two primary colors in a certain proportion. The mixed color replacement equation is established based on the gray balance. The mixed color replacement equation obtained is the result of the mixed color being completely replaced. The replacement rate is set according to the original to obtain the color separation result of the mixed color replacement; for the color parts where dark ink is used less, the corresponding light ink is used to obtain a better color reproduction effect, and the value relationship between light ink and dark ink is established, that is, when the use of dark ink is lower than the critical domain, light ink is used to replace it, and when it is higher than the critical domain, dark ink is used. The color separation range is expanded to the entire ten-color channel of the printer to obtain a better and more detailed color separation effect.

[0069] (3) This application is the research and development of printer color separation that has both expanded color gamut and improved color tone printing quality. Based on a ten-color inkjet printer, a color separation model is developed for a printer that combines the advantages of the two types of printers, and the color separation technology involved in the printer color control system is optimized. Based on color adaptive coding, the color separation model of a multi-color inkjet printer is self-encoded for the color gamut composed of the six dark inks C, M, Y, K, O, and G of the ten-color printer. A color separation model is established to obtain the color separation values ​​of each partition. The mapping relationship between dark ink and light ink is used to complete the conversion of the six-color color separation values ​​to the ten-color color separation values, and the entire color separation modeling of the ten-color channel of the printer is completed. A multi-color item fitting algorithm is used to construct a color separation model in each sub-partition; referring to the mid-mixed color replacement process (non-color structure process), mid-mixed color local replacement and mixed color replacement are developed, and their effects on the multi-color printer color separation model are determined; the light ink replacement method is developed to determine its influence on the multi-color printer color separation model, and a high-precision and high-quality color separation model of a multi-color printer with color adaptive coding is established. BRIEF DESCRIPTION OF THE DRAWINGS

[0070] Figure 1 It is a diagram of the proportional coefficients, density values ​​and dot area ratios of each level of the three-color scale when gray is balanced.

[0071] Figure 2 It is a multi-color fitting relationship diagram of C, M, Y and K.

[0072] Figure 3 It is a diagram showing the ratio coefficients of base colors at each level of the ladder scale, density values ​​and dot area ratios when O color is substituted.

[0073] Figure 4 It is a schematic diagram of the building surface vector data of the existing data in the area.

[0074] Figure 5 This is a schematic diagram of the smoothed DSM and slope data.

[0075] Figure 6 It is the color error diagram of the test samples in each partition after intermediate color substitution.

[0076] Figure 7 It is a color difference data graph obtained by calculating the error between the chromaticity value and the standard chromaticity value. DETAILED DESCRIPTION

[0077] The following, in conjunction with the accompanying drawings, further describes the technical solution of the high-precision color separation model of the color adaptive encoding multi-color printer provided by the present application, so that those skilled in the art can better understand the present application and implement it.

[0078] Compared with traditional four-color printers, multi-color printers have a larger color gamut and improved printing quality, which can meet the output requirements of high-fidelity images. At present, the research and development of color separation technology for printers is mainly focused on CMYKcm printers that increase the output level of printers and CMYKOG printers that increase the output color gamut of printers, that is, multi-color printers that increase light ink and multi-color printers that increase base colors.

[0079] For multi-color printers that add light ink, a replacement critical domain between light ink and dark ink is established on the basis of the four-color separation mechanism to achieve color separation on the entire printer channel, self-encode the printer color gamut, construct empirical models in each sub-color gamut to establish color separation models for each partition, and use mid-mixed color local replacement and mixed color replacement to complete printer color separation.

[0080] This application relies on a ten-color printer containing four light ink channels and two spot color channels to self-encode the color gamut composed of the six dark inks of the printer, and uses multivariate color gamut regression empirical analysis to establish three-color color separation models for each partition. According to the gray balance equation establishment method, the relationship between the local displacement of gray components and the local displacement of mixed colors is established. Finally, according to the density correspondence between light ink and dark ink, the displacement critical domain between light ink and dark ink is established to establish the color separation model of the printer within the ten-color channel; this application also randomly selects test samples from the IT8.7 / 4 sample set generated by the printer standard characteristic file to test and evaluate the color separation model.

[0081] Color separation model for multi-color inkjet printers based on color adaptive coding: self-encode the color gamut composed of the six dark inks C, M, Y, K, O, and G of a ten-color printer, establish a color separation model to obtain the color separation values ​​of each separation, and use the mapping relationship between dark ink and light ink to complete the conversion of the six-color separation value to the ten-color separation value, thus completing the entire color separation modeling of the printer's ten-color channels.

[0082] 1. Sample color adaptive encoding preparation

[0083] Calculate the hue angles of the six dark inks of the printer, and adaptively encode the printer colors into five areas: YOK, OMK, CMK, GCK, and GYK. Take samples in each color separation area to form a sample set for modeling each partition. In the GCK partition, G, C, and K are overprinted in a step of 10% within the range of 0-100% dot area rate to form a sample set for each partition. Turn off the printer color management, and directly print out the sample colors of the five partitions through the printer channel control.

[0084] 2. Multivariate color gamut regression analysis to establish color model for each division

[0085] (I) Multivariate color gamut regression analysis

[0086] The conversion model of the color separation model from the CIELAB color space to the CMYK color space belongs to multivariate multiple nonlinear regression. The relationship between the original color space and the target color space is established through multi-color terms. The coefficients are obtained using the least squares method to determine the mathematical conversion model of the two color spaces. Finally, the target color space and mixed color are calculated by the multi-color term relationship, and the decanomial multi-color terms are used to establish the color separation model for each separation.

[0087] The conversion model from CIELAB color space to CMYK color space in this application is a multivariate multiple nonlinear regression problem. Based on the correlation between multiple dependent variables and multiple independent variables, the following mathematical model is established for the multiple univariate linear regression problem: Assume that there are n independent variables x 1至n , corresponding to m dependent variables y 1至m , establish the relationship between the two sets of data:

[0088]

[0089] where β ij (i=0-n;j=0-m) are unknown parameters, ε j (j=0-m) is a random error, which is not independent of each other. Assuming that these error values ​​obey a multivariate normal distribution, the above model can be written in the form of a matrix as follows:

[0090]

[0091] The least square method is used to solve the coefficients of the multiple univariate regression equation and establish the regression relationship:

[0092] (X 11 , x 12 , …, x 1n ,;y 11 ,y 12 , …, y 1f )

[0093] (x 21 , x 22 , …, x 2n ,;y 21 ,y 22 , …, y 2f )

[0094] ……Formula 3

[0095] The multivariate multiple linear regression model is expressed as:

[0096]

[0097] ε is the error matrix. The error value is neglected and the coefficient matrix is ​​solved by the least square method to obtain the multivariate multiple linear regression equation. The color separation model of the printer is converted from the CIELab color space to the CMY color space. The dependent variable is the dot area rate of the three colors of yellow, magenta and cyan. The independent variables are L, a, b and the product terms formed by each other. Multivariate multiple nonlinear regression analysis is used to obtain the values ​​of other independent variables when the values ​​of the independent variables L, a, and b are known. The nonlinear regression problem is converted into a linear regression problem to simplify the calculation process.

[0098] (II) Model Building

[0099] Print out the sample sets of each partition, calculate the Lab value of the sample color as the basic data for building the model, and establish a multi-color regression equation with the Lab value as the independent variable and the dot area rate of the primary color of each partition as the dependent variable. Based on the multivariate color gamut regression analysis and GCK partitioning, the established multi-color regression equation is as follows:

[0100]

[0101] G, C, K represent the dot area ratio of the three primary colors in the partition, σ ij (i=1-3; j=0-9) represents the multi-color coefficient to be calculated. For all sample colors in the partition, equations are established. Let P 1 =L,P 2 =A,P 3 =B,P 4 =L 2 , P 5 =A 2 , P 6 =B 2 , P 7 =LA,P 8 =AB,P 9 =LB, transform the multivariate nonlinear regression problem into a multivariate linear regression problem, and obtain the matrix equation based on all sample colors as shown in Formula 6:

[0102]

[0103] Calculate σ by multiple linear regression analysis ij The least squares estimate of the partition is finally obtained, and the mathematical conversion model from the chromaticity value of the partition to the dot area ratio value is obtained. The multi-color regression color separation model is established for YOK, OMK, CMK, and GYK partitions in the same way. The regression model established for KCG partition is as shown in Formula 7:

[0104]

[0105] 3. Mixing colors, partial replacement and then color separation

[0106] The medium mixed color is partially replaced according to the original requirements. First, a gray balance curve is established for the CMYK4 colors, and then the medium mixed color replacement rate μ k Perform local replacement;

[0107] Calculate the density of each color block under the R filter, G filter, and B filter, and calculate the average value multiple times. First, completely replace the mixed color in the middle to determine the relationship between K and CMY.

[0108] D YB , D MB , D CB They are the densities of the three color inks Y, M, and C measured under the blue filter, D YG , D MG , D CG They are the density of Y, M, and C inks measured under the green filter, D YR , D MR , D CR are the densities of the three color inks Y, M, and C measured under the red filter, D k It is the visual density of monochrome black ink. It is calculated according to the equivalent neutral density calculation method to obtain the proportional coefficient of yellow, magenta and cyan inks when the neutral gray balance is achieved at each level; D e To achieve gray balance, the gray density value under each color filter is used to calculate the proportional coefficient of each level of the ladder. The proportional coefficient of each level is multiplied by the corresponding main density to obtain the density values ​​of cyan, magenta, and yellow required to form each level of neutral gray balance. The dot area ratio of each color block is calculated, and the empirical correction index is taken as 2. When gray balance is achieved, the dot area ratio of each level is as follows: Figure 1 shown.

[0109] Figure 2 This is the multi-color item fitting relationship diagram of C, M, Y and K. When the number of multi-color items is 3, a good fitting degree has been achieved. The equivalent substitution relationship between K and CMY is obtained:

[0110]

[0111] The displacement relationship is the result of replacing all mixed colors with black ink. The local displacement rate of mixed colors in the local displacement setting is μ k , the color separation value after all the mixed colors are replaced is C O , G O , K O , the K value of the KCG partition to be replaced is K O *μ k , and get the equivalent C K , M K ,Y K Color separation value, the number of color separations becomes CMYGK5 colors, the color separation value is C e 、M e , Y e, G e , K e , after the black ink in the GCK partition is partially replaced:

[0112]

[0113] The final color separation result is shown in Formula 9.

[0114] 4. Mixing, replacing and then separating colors

[0115] The orange O and green G of the ten-color printer are medium mixed colors. Based on the method of establishing the gray balance curve, the orange O is overprinted with yellow Y and magenta M in appropriate proportions. If it can produce the same color as the monochrome orange O, the density of the monochrome orange O at this time is calculated. This density is the density of the orange O equivalent to MY at the dot area rate at this time. The density is converted into dot area rate to establish the color replacement relationship between O and MY: first, design the monochrome gradient measurement and control strips of O, M, and Y, with a step unit of 5%. There are 20 color blocks with dot area rates from 0% to 100%, and measure the density of each sampling point in the measurement and control strip under the G filter and the B filter respectively.

[0116] According to the calculated density values ​​of each sampling point, the following equations are obtained between O and MY:

[0117]

[0118] Where D OG , D OB are the main density values ​​of color O measured under the green and blue filters, ω Y ,ω m D is the ratio coefficient of M and Y when forming the corresponding tone density of O. YG , D MG Refers to the density value of M and Y under the green filter; DYB and DMB refer to the density value of M and Y under the blue filter. The solution of the proportional coefficient is shown in formula 11:

[0119]

[0120] Obtain the M and Y ratio coefficients required to form each level of O color, and multiply them with the main density to obtain the actual required M and Y density as shown in formula 12:

[0121]

[0122] Figure 3 Middle, D Ye , D Me The actual effective density values ​​of Y and M required to form the color mixing O, F Y , F MIn order to obtain the effective dot area rate based on the effective density value using the Yule-Nielsen formula, a univariate multivariate color gamut regression analysis is used to determine the coefficients of the multicolor items and obtain the equivalent substitution relationship between O and MY as shown in Formula 13:

[0123]

[0124] The equivalent substitution relationship between G and CY is obtained as follows:

[0125]

[0126] Formula 13 and Formula 14 are equivalent color replacement values ​​when the 0 color and the G color are replaced to the maximum extent. The mixed color replacement rate is determined according to the original manuscript. The replacement rate when the O color is partially replaced by the M and Y colors is α, and the replacement rate when the G color is replaced by the C and Y colors is β, as shown in Formula 15 and Formula 16:

[0127]

[0128] Similarly, based on the KCG partitioning and the local replacement of the mixed color, the local replacement of the mixed color is performed again, and the final color separation result is formula 17:

[0129]

[0130] The local displacement rate of mixed color in the local displacement setting is μ k , the color separation value after all the mixed colors are replaced is C O , G O , K O , the K value of the KCG partition to be replaced is K O *μ k , and get the equivalent C K , M K ,Y K Color separation value, the number of color separations becomes CMYGK5 colors, the color separation value is C e 、M e , Y e , G e , K e .

[0131] 5. Light ink replacement method for color separation

[0132] The light ink and dark ink have the same tone, and a linear relationship is established between the two. C ,I M ,I K ,I LK ,I LLK ,I LM ,I LCThe dot area ratios of magenta, cyan, black, light black, light light black, light magenta, and light cyan are expressed by taking the density of the dark ink closest to the density when the light ink concentration is 100% as the basic data to establish the substitution relationship between the light ink and the dark ink;

[0133] According to the replacement relationship between light ink and dark ink, I K =aI LK ,I K =bI LLK ,I C =eI LC ,I M =fI LM Formula 18 is the linear mapping relationship between black ink and light black ink, and light-light black ink.

[0134]

[0135] a, b are the color separation critical domain of K. Print out the black and light ink gradation scale, calculate the density value of each color block, the density when the light black ink concentration is 100% is close to the density when the dark black ink concentration is 20%, let the critical domain of the two channels be 0.2, the value of b is 0.2, and the value of a is 0.45. When the black color separation value is less than the critical value of 0.2, let the black printing channel color separation value be 0, replace the black channel with light black, and calculate its color separation value k / 0.2 according to the linear transformation; if the black color separation value is greater than or equal to 0.2 and less than 0.45, let the black printing channel color separation value be 0, replace the black channel with light black, and calculate its color separation value I according to the linear transformation K / 0.45;

[0136] The substitution relationship between cyan and light cyan, and between magenta and light magenta are shown in formulas 4-18 and 4-19. The value of e is 0.3, and the separation value of f is 0.25. When the cyan separation value is greater than 0.3, the separation value of the cyan channel is set to 0, and replaced by the light cyan channel separation value Ic / 0.30; when the magenta separation value is greater than 0.25, the separation value of the magenta channel is set to 0, and replaced by the light magenta channel separation value IM / 0.25.

[0137]

[0138] According to the results of medium mixed color local replacement and mixed color local replacement, the light ink replacement is performed based on the critical domain of light ink replacement, and finally the color separation data within the ten color channels of the printer is obtained.

[0139] 6. Evaluation and Analysis of Color Separation Model for Multicolor Printers

[0140] According to the ICC characteristic file of the printer standard, the chromaticity values ​​of each color in the IT8.7 / 4 sample set are obtained. These chromaticity values ​​are used as known data to obtain the hue angle of each color, and compared with the hue angle range of the printer sub-color gamut. The sub-interval to which each color in the sample set belongs is confirmed. 60 samples are randomly selected from each partition sample as test samples. The multi-color regression equation of each interval is used to perform subspace color separation. On this basis, the medium mixed color local replacement color separation processing and mixed color local replacement processing are performed. Finally, the critical value of the replacement of dark ink and light ink is used to determine whether the light ink replacement color separation processing is required, and finally the color separation data within the entire printer color gamut is obtained. According to the color separation data, each channel of the printer is directly controlled to output the test sample, and the chromaticity coordinates of each color block in the sampling diagram are calculated to perform color difference calculation with the chromaticity values ​​in the standard characteristic file to evaluate the model.

[0141] (I) Inspection samples

[0142] The color gamut of a ten-color printer is encoded into five sub-areas: YOK, OMK, CMK, GCK, and GYK. Any color is obtained by mixing the two basic colors that make up the color gamut to which it belongs and black. The hue of the color is determined by the mixing ratio of the two basic colors, and the amount of black ink determines the brightness of the color. The hue angle of the color in each sub-gamut must be within the hue angle formed by the two basic hues. The hue angle is calculated, and the sub-gamut attribution is self-coded. According to the distribution in each partition, 60 samples are randomly selected from each partition, totaling 300 samples to form the test sample.

[0143] (II) Multivariate color gamut regression analysis color separation

[0144] The chromaticity values ​​of the 60 test samples in each partition are used as the input values ​​of the multi-color regression equation of each partition to obtain the three-color separation values ​​of each partition, print them out, and use the CIE1976L*a*b color difference formula to perform color difference calculation with the standard value. Figure 4 It is the color separation error after regression color separation of the test samples in each partition.

[0145] After fitting the multi-color items of each partition, the error value of the test sample is relatively large, but it is all within the range of the hue angle of each partition.

[0146] (III) Partial replacement of mixed colors

[0147] The three-color separation data of each color block is obtained according to the multi-color regression model established in each partition. The gray component is partially replaced and the replacement rates are set to 0.2, 0.4, 0.6, and 0.8 respectively. The color separation data under each replacement rate is obtained and printed out. The chromaticity value is calculated and the color difference operation is performed with the chromaticity value of the standard characteristic file using the CIE1976L*a*b* color difference formula. The results are as follows: Figure 5 shown.

[0148] When the medium color mixing replacement rate is 0.2, the color difference of the test samples in each partition is smaller than that in other replacement rates. The color separation data when the medium color mixing replacement rate of each partition is 0.2 are selected to enter the next step of testing.

[0149] (IV) Color mixing and replacement

[0150] After determining the mixed color replacement, use mixed color replacement on each partition test sample, analyze the color difference between using mixed color replacement and not using mixed color replacement relative to the standard chromaticity value, set the replacement rate to 0.2, 0.4, 0.6, and 0.8 respectively, obtain the color separation data under each group replacement rate, print it out, calculate its chromaticity value, and calculate the color difference, such as Figure 6 shown.

[0151] A relatively good effect was achieved when the color mixing replacement rate was 0.2, so the test sample data of each partition when the color mixing replacement rate was 0.2 was taken for the next step of verification.

[0152] (V) Light ink replacement rate and color separation

[0153] According to the determined light ink replacement critical value, when the determined medium color replacement rate and color replacement rate are 0.2, the color separation of each partition test sample is performed again. The color separation value of each partition test sample in the entire printer's ten color channels is obtained, and the color value is calculated and compared with the standard color value for error calculation to obtain the following: Figure 7 The color difference data shown in the table also shows the overall color difference of the test samples in the last column.

[0154] The color separation error of the model established in the present application is smaller than the standard chromaticity value error, and the medium mixed color local replacement, mixed color replacement and light ink replacement have obvious effects on reducing the model error.

Claims

1. Color adaptive coding multi-color printer high-precision color separation model, characterized by: For multi-color printers that add light ink, a replacement critical domain between light ink and dark ink is established on the basis of the four-color separation mechanism to achieve color separation on the entire printer channel, self-encode the printer color gamut, construct empirical models in each sub-gamut to establish color separation models for each partition, and use mid-mixed color local replacement and mixed color replacement to complete printer color separation; Color separation model for multi-color inkjet printers based on color adaptive coding: self-encode the color gamut composed of the six dark inks C, M, Y, K, O, and G of a ten-color printer, establish a color separation model to obtain the color separation values ​​of each separation area, and use the mapping relationship between dark ink and light ink to complete the conversion of the six-color separation value to the ten-color separation value, thus completing the entire color separation modeling of the ten-color channel of the printer; The process of establishing the color separation model of a multi-color printer based on color adaptive coding is as follows: after verifying that the chromaticity of light ink is consistent with that of dark ink, the light ink channel of the ten-color printer is separated, and the color gamut of the printer composed of six dark inks is color adaptively encoded according to color adaptive coding, and self-encoded into five subspaces. In each subspace, a subspace color separation model is established using a multi-color term regression equation, a gray balance equation is established, and a local replacement rate of medium mixed colors is set to obtain a color separation result of local replacement of medium mixed colors. The mixed color part of the original is mixed by two primary colors in a certain proportion. A mixed color replacement equation is established based on gray balance. The obtained mixed color replacement equation is the result of complete replacement of the mixed colors. The replacement rate is set according to the original to obtain a color separation result of mixed color replacement. For the color part with less dark ink, the corresponding light ink is used to obtain a better color reproduction effect, and a value relationship between light ink and dark ink is established, that is, when the use of dark ink is lower than the critical domain, light ink is used to replace it, and when it is higher than the critical domain, dark ink is used, so that the color separation range is expanded to the entire ten-color channel of the printer to obtain a better color separation effect.

2. According to the color adaptive coding multi-color printer high-precision color separation model of claim 1, it is characterized in that: Based on color adaptive coding, a multi-color item fitting method is used in each sub-color gamut of a multi-color printer to establish a color separation model for each partition, and on this basis, the medium mixed color local displacement, mixed color displacement and light ink displacement are improved to optimize the model: first, in each sub-color gamut of a multi-color printer, a sampling sample is established, and a three-color color separation model for each partition is constructed by a multi-color item fitting method; second, based on the establishment method of the gray balance equation, a medium mixed color local displacement relationship and a mixed color displacement relationship are established, and the medium mixed color local displacement equation and the mixed color displacement equation are used to further separate the color separation results of each partition to optimize the model; third, a critical domain relationship between light ink and dark ink is established. When the dark ink usage is lower than the critical value, light ink is used for displacement printing; when the dark ink usage is higher than the critical domain, dark ink is used directly.

3. According to the color adaptive coding multi-color printer high-precision color separation model of claim 1, it is characterized in that: Preparation for sample color adaptive encoding: calculate the hue angles of the six dark inks of the printer, and adaptively encode the printer colors into five areas: YOK, OMK, CMK, GCK, and GYK. Samples are taken in each color separation area to form modeling sample sets for each partition. In the GCK partition, G, C, and K are overprinted in the range of 0-100% dot area rate with a step size of 10% to form sample sets for each partition. Turn off the printer color management, and directly print out the sample colors of the five partitions through the printer channel control.

4. The high-precision color separation model of the color adaptive coding multi-color printer according to claim 1 is characterized in that: Multivariate color gamut regression analysis: The conversion model of the color separation model from the CIELAB color space to the CMYK color space belongs to multivariate multiple nonlinear regression. The relationship between the original color space and the target color space is established through multi-color terms. The coefficients are obtained using the least squares method to determine the mathematical conversion model of the two color spaces. Finally, the target color space and mixed color are calculated by the multi-color term relationship, and the decanate multi-color terms are used to establish the color separation model for each partition.

5. The color adaptive coding multi-color printer high-precision color separation model according to claim 4 is characterized in that: The conversion model from CIELAB color space to CMYK color space is a multivariate multiple nonlinear regression problem. Based on the correlation between multiple dependent variables and multiple independent variables, the following mathematical model is established for the multiple univariate linear regression problem: Assume that there are n independent variables x 1至n , corresponding to m dependent variables y 1至m , establish the relationship between the two sets of data: where β ij (i=0-n;j=0-m) are unknown parameters, ε j (j=0-m) is a random error, which is not independent of each other. Assuming that these error values ​​obey a multivariate normal distribution, the above model can be written in the form of a matrix as follows: The least square method is used to solve the coefficients of the multiple univariate regression equation and establish the regression relationship: The multivariate multiple linear regression model is expressed as: ε is the error matrix. The error value is neglected and the coefficient matrix is ​​solved by the least square method to obtain the multivariate multiple linear regression equation. The color separation model of the printer is converted from the CIELab color space to the CMY color space. The dependent variable is the dot area rate of the three colors of yellow, magenta and cyan. The independent variables are L, a, b and the product terms formed by each other. Multivariate multiple nonlinear regression analysis is used to obtain the values ​​of other independent variables when the values ​​of the independent variables L, a, and b are known. The nonlinear regression problem is converted into a linear regression problem to simplify the calculation process.

6. According to the color adaptive coding multi-color printer high-precision color separation model of claim 4, it is characterized in that Model: Print out the sample sets of each partition, calculate the Lab value of the sample color as the basic data for building the model, use the Lab value as the independent variable and the dot area rate of the base color of each partition as the dependent variable to establish a multi-color regression equation. Based on the multivariate color gamut regression analysis and GCK partition, the established multi-color regression equation is as follows: G, C, K represent the dot area ratio of the three primary colors in the partition, σ ij (i=1-3; j=0-9) represents the multi-color coefficients to be determined. For all sample colors in the partition, equations are established, and P1=L, P2=A, P3=B, P4=L 2 , P5=A 2 , P6=B 2 , P7=LA, P8=AB, P9=LB, transform the multivariate nonlinear regression problem into a multivariate linear regression problem, and obtain the matrix equation based on all sample colors as shown in Formula 6: Calculate σ by multiple linear regression analysis ij The least squares estimate of the partition is finally obtained, and the mathematical conversion model from the chromaticity value of the partition to the dot area ratio value is obtained. The multi-color regression color separation model is established for YOK, OMK, CMK, and GYK partitions in the same way. The regression model established for KCG partition is as shown in Formula 7:

7. The color adaptive coding multi-color printer high-precision color separation model according to claim 1, characterized in that: Partial replacement of medium mixed colors and then color separation: The medium mixed colors are partially replaced according to the original requirements. First, a gray balance curve is established for the CMYK4 colors, and then the medium mixed color replacement rate μ k Perform local replacement; Calculate the density of each color block under the R filter, G filter, and B filter respectively, calculate multiple times to get the average value, first completely replace the mixed color in the middle to determine the relationship between K and CMY; D YB , D MB , D CB They are the densities of the three color inks Y, M, and C measured under the blue filter, D YG , D MG , D CG They are the density of Y, M, and C inks measured under the green filter, D YR , D MR , D CR are the densities of the three color inks Y, M, and C measured under the red filter, D k It is the visual density of monochrome black ink. It is calculated according to the equivalent neutral density calculation method to obtain the proportional coefficient of yellow, magenta and cyan inks when the neutral gray balance is achieved at each level; D e To achieve gray balance, the gray density value under each color filter is used to calculate the proportional coefficient of each level of the ladder scale, and the proportional coefficient of each level is multiplied by the corresponding main density to obtain the density values ​​of cyan, magenta, and yellow required to form each level of neutral gray balance; the dot area rate of each color block is calculated, and the empirical correction index is taken as 2; When the number of multi-color items is 3, a good fit is achieved, and the equivalent substitution relationship between K and CMY is obtained: The displacement relationship is the result of replacing all mixed colors with black ink. The local displacement rate of mixed colors in the local displacement setting is μ k , the color separation value after all the mixed colors are replaced is C O , G O , K O , the K value of the KCG partition to be replaced is K O *μ k , and get the equivalent C K , M K ,Y K Color separation value, the number of color separations becomes CMYGK5 colors, the color separation value is C e 、M e , Y e , G e , K e , after the black ink in the GCK partition is partially replaced: The final color separation result is shown in Formula 9.

8. The color adaptive coding multi-color printer high-precision color separation model according to claim 1, characterized in that: Mixed color replacement and color separation: Orange O and green G are medium mixed colors of a ten-color printer. Based on the gray balance curve establishment method, the orange O is overprinted with yellow Y and magenta M in appropriate proportions. If it can produce the same color as the single-color orange O, the density of the single-color orange O is calculated. This density is the density of the orange O equivalent to MY at the dot area rate at this time. The density is converted into the dot area rate to establish the color replacement relationship between O and MY: First, design the single-color gradient measurement and control strips of O, M, and Y, with a step unit of 5%, and a total of 20 color blocks with dot area rates ranging from 0% to 100%. Measure the density of each sampling point in the measurement and control strip under the G filter and the B filter respectively; According to the calculated density values ​​of each sampling point, the following equations are obtained between O and MY: Where D OG , D OB are the main density values ​​of color O measured under the green and blue filters, ω Y ,ω m D is the ratio coefficient of M and Y when forming the corresponding tone density of O. YG , D MG Refers to the density value of M and Y under the green filter; DYB and DMB refer to the density value of M and Y under the blue filter. The solution of the proportional coefficient is shown in formula 11: Obtain the M and Y ratio coefficients required to form each level of O color, and multiply them with the main density to obtain the actual required M and Y density as shown in formula 12: D Ye , D Me The actual effective density values ​​of Y and M required to form the color mixing O, F Y , F M In order to obtain the effective dot area rate based on the effective density value using the Yule-Nielsen formula, a univariate multivariate color gamut regression analysis is used to determine the coefficients of the multicolor items and obtain the equivalent substitution relationship between O and MY as shown in Formula 13: The equivalent substitution relationship between G and CY is obtained as follows: Formula 13 and Formula 14 are equivalent color replacement values ​​when the 0 color and the G color are replaced to the maximum extent. The mixed color replacement rate is determined according to the original manuscript. The replacement rate when the O color is partially replaced by the M and Y colors is α, and the replacement rate when the G color is replaced by the C and Y colors is β, as shown in Formula 15 and Formula 16: Similarly, based on the KCG partitioning and the local replacement of the mixed color, the local replacement of the mixed color is performed again, and the final color separation result is formula 17: The local displacement rate of mixed color in the local displacement setting is μ k , the color separation value after all the mixed colors are replaced is C O , G O , K O , the K value of the KCG partition to be replaced is K O *μ k , and get the equivalent C K , M K ,Y K Color separation value, the number of color separations becomes CMYGK5 colors, the color separation value is C e 、M e , Y e , G e , K e .

9. The color adaptive coding multi-color printer high-precision color separation model according to claim 1, characterized in that: Light ink replacement method for color separation: light ink and dark ink have the same tone, and a linear relationship is established between the two. C ,I M ,I K ,I LK ,I LLK ,I LM ,I LC The dot area ratios of magenta, cyan, black, light black, light light black, light magenta, and light cyan are expressed by taking the density of the dark ink closest to the density when the light ink concentration is 100% as the basic data to establish the substitution relationship between the light ink and the dark ink; According to the replacement relationship between light ink and dark ink, I K =aI LK ,I K =bI LLK ,I C =eI LC ,I M =fI LM The linear mapping relationship of , formula 18 is the replacement relationship between black ink and light black ink, light light black ink; a, b are the color separation critical domain of K. Print out the black and light ink gradation scale, calculate the density value of each color block, the density when the light black ink concentration is 100% is close to the density when the dark black ink concentration is 20%, let the critical domain of the two channels be 0.2, the value of b is 0.2, and the value of a is 0.

45. When the black color separation value is less than the critical value of 0.2, let the black printing channel color separation value be 0, replace the black channel with light black, and calculate its color separation value k / 0.2 according to the linear transformation; if the black color separation value is greater than or equal to 0.2 and less than 0.45, let the black printing channel color separation value be 0, replace the black channel with light black, and calculate its color separation value I according to the linear transformation K / 0.45; The substitution relationship between cyan and light cyan, and between magenta and light magenta are shown in formulas 4-18 and 4-19. The value of e is 0.3, and the color separation value of f is 0.

25. When the cyan color separation value is greater than 0.3, the color separation value of the cyan channel is set to 0, and replaced by the color separation value of the light cyan channel Ic / 0.30; when the magenta color separation value is greater than 0.25, the color separation value of the magenta channel is set to 0, and replaced by the color separation value of the light magenta channel IM / 0.

25. According to the results of medium mixed color local replacement and mixed color local replacement, the light ink replacement is performed based on the critical domain of light ink replacement, and finally the color separation data within the ten color channels of the printer is obtained.

Images