基于BP神经网络的CRT显示器色彩管理新模型  

New Color Management Model for CRT Monitor Based on Neural Network

作　　者：黎新伍[1] 

机构地区：[1]江西财经大学电子商务系,江西南昌330013

出　　处：《液晶与显示》2008年第2期235-240,共6页Chinese Journal of Liquid Crystals and Displays

基　　金：江西省教育厅重点资助项目(No.2007259);中国航空科学基金(No.02153071)

摘　　要：CRT显示器的色彩管理是信息光学中图像复制的关键技术之一。在分析呈色原理的基础上,提出了一种新的CRT显示器色彩管理模型。首先,采用标准色靶作为实验样本,并以彩色梯尺区色块代表整个色空间用以推导模型;其次,介绍了实验数据采集方法,并建立若干数据库用以推导模型;随后,利用Steffensen算法改进BP神经网络,并针对显示器色彩管理的实际运算,提出若干加速网络收敛的措施,克服了BP神经网络收敛速度慢的问题。实验结果表明本模型能够提高CRT显示器的色彩管理精度,工程实践中能够用于CRT显示器的色彩管理。Color management for the CRT monitor is one of the key techniques in the color image reproduction in information optics. A new color management model was presented based on analyzing the CRT color rendering principle. First, the paper took standard color target for experimental sample and substitutes color blocks in color shade district for complete color space. Second, data collecting method was introduced and some data bases for deducing the model were created. Then, Steffensen algorithm was used to correct BP neural network and some measures were presented to speed up convergence of the BP model. Finally the experimental results showed that the model can improve CRT monitor color management accuracy and can be used in its color management practically.

关 键 词：色彩管理 Steffensen算法 BP神经网络 标准色靶 

分 类 号：TN27[电子电信—物理电子学] TN941.1