A Study Of Color Transfer Feature Of Prepress System

Application of digital devices in field of prepress makes prepress system configuration more flexible and more convenience to prepress CMS(Color Management System). At the same time, some possible negative ef f ects-of-prepress CMS are brought. Complicated nonlinear relation is between in the devices input and output of prepress devices. In conventional CMS, the devices input and output relation is described by a linear model. However, noisy signal can't be processed in the linear model very well, so does some other possible factors.In this paper, on the basis of analyzing the present CMS and related models, a prototype CMS based on BP neural network theory is developed. The contents in this paper are as following:1. By analyzing CRT and printer working principle and combined with BP neural network theory, a printer calibration model is put forward by sorting experiment data according to hue angel range, and CRT and printer characteristic model is created. On the basis of device characteristic model, a gamut mapping algorithm is provided based on BP neural network theory.2. By discussing CMS theory and method, experiment is devised and carriedout and a great deal of data is obtained. On the color management prototype system developed on MatLab 6. 5 platform, experiment data is used to train the neural networks in the system, neural weights and biases are ascertained. CRT and printer characteristic model and printer calibration model are analyzed, validated and evaluated. The result indicates that the model precision is high enough that is less than the lest threshold of human perceived. The gamut mapping algorithm slashes the color errors between CRT and printer.3. Printer calibration model by sorting experiment data according to hue angel arrange has rather high precision, and can be used to describe other kinds of printing devices to instruct operators to adjust ink amount according to desired color.