During trial printing, digital printing color separation can adjust the threshold value according to the printing effect in real time to achieve the best threshold value and the best printing effect. As fabrics are woven from various fibers, their properties in warp and weft directions will be different. If not considered, unexpected stripes may appear in the printed image. To prevent this, we introduce two parameters to represent the difference between longitude and latitude, which are called longitude coefficient x and latitude coefficient y. In this way, we can adjust these two coefficients to adapt to the difference between longitude and latitude attributes. Different output devices will have different resolutions, and their color tone values will also be different. The worse equipment can represent fewer colors, while the better equipment will consider that the CMYK color model is actually a CMY color model, and black is only a supplement. Therefore, we first deal with the three colors of cyan, magenta, and yellow in the color separation process, and finally propose black from the three colors.

Definition of color separation process parameters: n pixel matrix dimension is used to determine the gray level of the output image, gray level=nn+1; x. Y warp coefficient and weft coefficient, representing the difference between the warp and weft properties of fabrics; F Threshold coefficient, which determines the threshold value of color separation together with the gray scale ratio; Space gray level ratio, the ratio of gray level before and after color separation; Error, the deviation of pixel color gray before and after color separation; The value of the color separation threshold is used to determine the number of color points of the output image. Input parameters: threshold coefficients of n pixel matrix dimensions: fc, fM, fY (0,1); x. Y longitudinal coefficient and zonal coefficient; Output result: the number of coloring points in four colors: Cv, Mv, Yv, Kv cyan, magenta, and black. Color separation processing steps: color model conversion Color model conversion is performed according to the following formula: C=255-R; M=255-G； Y=255-B Determine the number of coloring points of three colors without considering black temporarily. Take cyan as an example, set its color gray scale value as C. To determine the number of coloring points C0 of this color, solve it as follows.

Calculate the intermediate variable: error dispersion The error of a pixel after the second step of processing is error=C-C0space To ensure the image effect, this error needs to be processed. The specific method is to disperse and accumulate the error to the surrounding points in a certain proportion, so that it can be compensated at the adjacent points.

The filter algorithm diagram shows the pixel to be processed. The error value of this point after the second step of processing is error, and the error is dispersed according to the proportional coefficient in: add 8/42 of the error to the first image point on the right, 4/42 to the second image point on the right, 2/42 to the second image point on the right of the next line, and so on. Add the error dispersion to the 12 adjacent related points for compensation, This is the error allocation scheme of Stucki filtering algorithm. In addition, considering the difference of fabric warp and weft attributes, we modified the error dispersion coefficient of Stucki filtering algorithm. The specific method is to multiply the coefficient in the figure by the warp coefficient and the weft coefficient x and y, and take the product as the final error dispersion coefficient. The result of this can make the color separation result adapt to a variety of fabrics with different warp and weft attributes. The image processed by this method has a good output effect because it involves more points.