基于径向基函数神经网络的柔印专色配色系统  

Spot-color matching system of flexo printing based on radius basis function neural network

作　　者：杨秋娟[1] 周世生[1] 罗如柏[1] 

机构地区：[1]西安理工大学印刷包装工程学院,西安710048

出　　处：《计算机应用》2013年第6期1787-1789,共3页journal of Computer Applications

基　　金：陕西省自然科学基础研究项目(2011JQ8017);陕西省教育厅科学研究项目(11JK0872);西安理工大学高层次人才科研启动基金资助项目(104-211003);陕西省"13115"科技创新工程技术研究中心项目(2009ZDGC-06)

摘　　要：针对目前人工配色受配色者的生理、心理等主观因素影响,使产品质量难以保证这一问题,提出了基于径向基函数(RBF)神经网络的计算机柔印配色方法。该方法通过印刷实验获得样本数据,利用K均值聚类算法确定隐含层节点中心、采用伪逆法计算输出权值等参数,完成配色模型的建立。该配色模型可以快速完成柔印配色,并且具有较高的配色精度。In order to solve the problem that human physiology and psychology have enormous implications to the color matching artificially and it is difficult to guarantee the product quality, a computer color matching method of flexography based on RBF （Radius Basis Function） neural network was presented. The method obtained the sample data with experimental method, determined the center of hidden layer and output layer weights in RBF neural network by K- means clustering algorithm and pseudo-inverse technique and gained the spot-color matching system of flexo printing based on RBF neural network. The spot-color matching system can match the flexography color with high accuracy and high speed.

关 键 词：柔印 专色 配色 径向基函数 神经网络 

分 类 号：TP391.75[自动化与计算机技术—计算机应用技术]