一种基于视觉特性和灰度补偿改进的LDR图像增强方法  

An improved LDR image enhancement algorithm based on visual characteristics and gray scale compensation

作　　者：何保畅 耿修瑞[1,2] 胡文龙[1,2] HE Baochang;GENG Xiurui;HU Wenlong(Institute of Electronic, Chinese Academy of Sciences, Beijing 100190, China;Key Laboratory of Geo-spatial Information Processing and Application System Technology, Chinese Academy of Sciences, Beijing 100190, China;University of Chinese Academy of Sciences, Beijing 100049, China)

机构地区：[1]中国科学院电子学研究所,北京100190 [2]中国科学院空间信息处理与应用系统技术重点实验室,北京100190 [3]中国科学院大学,北京100049

出　　处：《中国科学院大学学报（中英文）》2018年第3期391-401,共11页Journal of University of Chinese Academy of Sciences

基　　金：国家自然科学基金(61331017)资助

摘　　要：针对传统空域增强算法的不足,LDR(layered difference representation)算法构造一种二维直方图,该直方图包含像素间上下文信息。基于此,提出一种基于LDR的改进算法。该算法将视觉特性引入直方图,构造二维视觉差异直方图,再利用该直方图和LDR进行初步增强。最后采用灰度补偿方法改善过增强现象。实验结果显示该算法可产生更好的视觉效果,并适当提高整体亮度和对比度,消除人工伪影,且无需调整参数,满足实时处理的要求。In view of the drawbacks of traditional spatial image enhancement algorithms,this paper proposes an improved algorithm based on the layered difference representation（LDR） algorithm which constructs a two-dimensional histogram containing the spatial context information. The histogram exploits the visual characteristics to construct a two-dimensional visual difference histogram,which is consequently used in the enhancement of the input images with LDR. Finally,the grayscale compensation method is used to ameliorate the over-enhancement problem.Experimental results show that the proposed algorithm produces better visual effects（performance）,improves the overall brightness and contrast,and eliminates the artifacts. The proposed method is free of parameter selection and meets the real-time processing requirement.

关 键 词：图像增强 二维视觉差异直方图 视觉特性 灰度补偿 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]