限制对比度的多层POSHE自适应图像增强算法  

Adaptive image enhancement algorithm based on contrast limited multilayered POSHE

作　　者：杨光[1] 吴钟建[1] 罗镇宝[1] 王浩宇[2] 张龙[1] 

机构地区：[1]西南技术物理研究所,四川成都610041 [2]西南自动化研究所,四川绵阳621000

出　　处：《激光与红外》2013年第1期85-89,共5页Laser & Infrared

摘　　要：部分子块重叠直方图均衡算法(POSHE)是一种常用的图像增强方法,由于该算法存在过增强和块效应问题,不利于目标的识别和跟踪。为解决该问题,提出限制对比度的多层POSHE图像增强方法(CLMPOSHE),通过对每层子块对比度自适应限制来消除过增强现象,然后采用多层POSHE增强处理,最后将每层改进的POSHE处理结果自适应加权融合。大量实验仿真表明,该方法能有效增强雾天图像。Partially subblock overlapped HE is an usual image enhancement method, it is difficult for target recogni tion and tracking because of overenhancement and block effect within POSHE. In order to solve the problem, a con trast limited muhilayered POSHE （CLMPOSHE） method is proposed, contrast of each layer subblocks is adaptively limited to eliminate overenhancement, then muhilayered POSHE processing is executed, finally each layer POSHE al gorithm is fused with adaptive weights. Simulation experiment results show that the proposed algorithm is effective in enhancing foggy images.

关 键 词：图像增强 CLMPOSHE 对比度限制 图像融合 自适应 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]