Application of Grey Relational Analysis and SVM for Solid Density Control in Offset Color Printing

Abstract

Abstract: Some appropriate print samples (according to the ISO standards of ink color values) were selected to set up a mathematical model which showed the relationship between density and color. The model was based on Support Vector Machine algorithm. Another evaluation system was established based on Gray Relation to decide the contribution of K, ΔL*, Δa* and Δb* to density. The final mean square error of the training sample and testing sample was 0.001 and 0.003 respectively. The evaluation system showed that the biggest impact on C，M，Y，K solid ink density was Δa* , ΔL*, K and Δb* respectively, the grey relational degree was 0.7587，0.8340, 0.7591 and 0.8711 respectively. The results of this study indicated that vector machine algorithm could accurately predict density values from chromaticity values and that the grey evaluation system of the solid ink density can be used to evaluate how big the influence of K，ΔL*，Δa*，Δb* on solid density are in offset printing.