首页 | 本学科首页   官方微博 | 高级检索  
     

基于鲁棒估计的Retinex图像增强方法
引用本文:王晓莉,文华,罗予频.基于鲁棒估计的Retinex图像增强方法[J].计算机应用,2006,26(7):1611-1613.
作者姓名:王晓莉  文华  罗予频
作者单位:1. 清华大学,自动化系,北京,100084
2. 长沙理工大学,设计艺术学院,湖南,长沙,510015
摘    要:针对原变分Retinex方法的人工伪影问题,指出理想重建光照应该具有分区域恒常的特性,提出一种以阴影边界作为出格点、基于鲁棒估计的变分Retinex图像增强方法。用合成图像与实际图像进行实验表明,本改进方法不仅可以有效压缩图像动态范围,显示淹没在阴影区域的细节,而且解决了原Retinex方法的伪影和扭曲等问题。

关 键 词:Retinex  鲁棒估计  变分方法  动态范围压缩
文章编号:1001-9081(2006)07-1611-03
收稿时间:2006-01-19
修稿时间:2006-01-192006-03-06

Retinex image enhancement method based on robust estimation
WANG Xiao-li,WEN Hua,LUO Yu-pin.Retinex image enhancement method based on robust estimation[J].journal of Computer Applications,2006,26(7):1611-1613.
Authors:WANG Xiao-li  WEN Hua  LUO Yu-pin
Affiliation:1. Department of Automation, Tsinghua University, Bering 100084, China; 2. Institute of Design Art, Changsha University of Science and Technology, Changsha Hunan 510015, China
Abstract:Because of failing to reconstruct piecewise constant illumination, variation Retinex method suffers from artificial halos and color distortion. A Retinex method based on robust estimation was proposed, which took shadow edges as outliers and smoothed range by range. Application of method to synthetic image and natural image proves that the improved method preserves high dynamic range compression and avoids halo effect and color distortion.
Keywords:Retinex  robust estimation  variation method  dynamic range compression
本文献已被 CNKI 维普 万方数据 等数据库收录!
点击此处可从《计算机应用》浏览原始摘要信息
点击此处可从《计算机应用》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