Research Of Printing Medium Color Consistency

With the development of computer technology and network technology , the color information which is digitized can be described and transfered conveniently. Then the need for color quality of printing and productivity effect has become more critical. Meanwhile, increasingly high requirement of the color quality and production efficiency of printed products has been put forward. Consequently,virtual proofing comes into being in the full digitalization process of printing.This article mainly expounds and validates BP neural network algorithm to achieve method and effect of color space conversion in virtual proofing process. (1) Design the virtual proofing process scheme.Firstly, test the screen quality including the level characteristic, contrast,display consistency, stability and so on to evaluate the screen quality. And set the parameter such as the color temperature and gamma value to calibrate and characterize the monitor. Then test ability of monitor simulating four colors printing and gain the testing data finally. (2) Based on the comparison of the different training methods, transforming functions, several hidden layer number and the neuron number of hidden layer, this paper gain a BP neural network model of transforming CMYK to Lab color space, using the above-mentioned testing data.Finally, evaluate network effect in the color space transformation through linear regression analysis.This neural network model provides a new thought of color space transformation arithmetic for color management software corporation. And it has a great instructional effect on the virtual proofing theory and mean great values to the practicality.