首页 | 本学科首页   官方微博 | 高级检索  
     

基于样本空间学习算法的彩色匹配方法研究
引用本文:苏晓红,王亚东,郭茂祖. 基于样本空间学习算法的彩色匹配方法研究[J]. 自动化学报, 2001, 27(2): 186-193
作者姓名:苏晓红  王亚东  郭茂祖
作者单位:1.哈尔滨工业大学计算机科学与工程系,哈尔滨;
基金项目:国家自然科学基金! ( 699750 0 5),国际合作项目彩色匹配系统资金资助项目
摘    要:提出基于样本空间学习箅法的RGB(Red,Green,Blue)→CMYK(Cyan,Magenta,Yellow,blacK)颜色空间变换方法,仅需测试400组纯色色块,即可在不同的打印(抖动)模式、墨水和纸张条件下,对RGB→CMYK转换规律进行有效的学习.该算法学习能力强、打印速度快,通过在模型中增加高次项,提高了彩色匹配的精度,进一步解决了彩色匹配的自动化和通用性问题.同时,借助于CIELab均匀颜色空间对图像进行动态范围调整,还起到了减小色偏、突出图像细节、增强图像对比度的效果.

关 键 词:彩色匹配   样本空间   机器学习   多项式回归   色差
收稿时间:1999-07-19
修稿时间:1999-07-19

Research of Color Matching Method Based on Learning Algorithm in Version Space
SU Xiao-hong,WANG Ya-Dong,GUO Mao-Zu. Research of Color Matching Method Based on Learning Algorithm in Version Space[J]. Acta Automatica Sinica, 2001, 27(2): 186-193
Authors:SU Xiao-hong  WANG Ya-Dong  GUO Mao-Zu
Affiliation:1.Department of Computer Science and Engineering,Harbin Institute of Technology,Harbin;Canotec Corporation Incorporated,Japan
Abstract:A color space transformation method from RGB to CMYK based on knowledge discovery is presented in this paper. The RGB→CMYK for different color ink-jet printers and paper can be found out after only four hundred pure color blocks are tested. This method has a good learning ability and fast printing speed. By adding high-order item to the model, the precision of color matching is improved, and at the same time the problem of automatic color matching and generalization is solved. With the aid of CIELAB uniform color space, the RGB→CMYK laws are well adjusted to stress the image details and enhance image contrast.
Keywords:Color matching   version space   machine learning   multiple polynomial regression   color difference.
本文献已被 CNKI 维普 万方数据 等数据库收录!
点击此处可从《自动化学报》浏览原始摘要信息
点击此处可从《自动化学报》下载全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号