一种直方图均衡插值的图像细节增强方法  

Method of Image Detail Enhancement Using Histogram Equalization and Interpolation

作　　者：徐建东[1] XU Jiandong(Jiangsu Guoguang Information Industry Co.,Ltd.,Changzhou 213001,China)

机构地区：[1]江苏国光信息产业股份有限公司,江苏常州213001

出　　处：《江苏理工学院学报》2020年第2期23-29,共7页Journal of Jiangsu University of Technology

基　　金：江苏省现代教育技术研究2019年度智慧校园专项课题“大数据背景下稀疏向量学习算法理论与应用研究”(2019-R-75637)。

摘　　要：传统直方图均衡具有灰度级减少、细节丢失和过度增强等不足,为此,提出了一种基于直方图均衡插值的图像细节增强方法。首先,算法对输入图像作直方图均衡处理;其次,在传统直方图均衡的直方图相邻灰度间隔值从大到小的位置,插入某一灰度SP构成新直方图;最后,将新直方图的灰度值按照从小到大的顺序一一映射至原图像直方图中,并输出增强图像。与其它算法比较,信息熵指标始终排名第1,表明了直方图均衡插值算法在图像细节保留方面的优越性;同时,算法增强的图像视觉效果清晰、柔和。Traditional histogram equalization has some defects,such as gray level reduction,over enhancement and detail loss.To solve these problems,this paper proposes a method of image detail enhancement based on histogram equalization interpolation.Firstly,the algorithm makes histogram equalization for the input image;Secondly,in the middle of the adjacent gray levels of the equalization histogram,a gray SP is inserted to form the new histogram according to the rules from large to small;Finally,the gray value of the new histogram is mapped to the original image histogram according to the order from small to large,and the enhanced image is output at the same time.Compared with other algorithms,the information entropy index of this algorithm always ranks first,which shows the superiority of histogram equalization interpolation algorithm in image visual effect detail preservation.At the same time,the image enhanced by this algorithm is clear and soft.

关 键 词：直方图均衡 图像增强 细节保留 灰度插值 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]