基于局部熵-局部对比度和双区域直方图均衡化的红外图像增强  

Infrared Image Enhancement Based on Local Entropy-Local Contrast and Dual-area Histogram Equalization

作　　者：何智博 曾祥进[1] 邓晨 宋彭彭 HE Zhibo;ZENG Xiangjin;DENG Chen;SONG Pengpeng(School of Computer Science and Engineering,Wuhan Institute of Technology,Wuhan 430205,China)

机构地区：[1]武汉工程大学计算机科学与工程学院,湖北武汉430205

出　　处：《红外技术》2023年第6期598-604,共7页Infrared Technology

基　　金：国家自然科学基金项目(61502354);湖北省教育厅重点研究项目(D20171503)。

摘　　要：为了改善红外图像的视觉效果,突出细节信息,同时抑制噪声。提出了结合改进的LC显著性检测和双区域直方图均衡化的红外图像增强方法。首先使用结合局部熵加权的LC显著性检测算法得到显著图。然后使用K-means算法对显著图进行自适应分割得到前景区域和背景区域。最后对前景区域进行结合局部方差的改进直方图均衡化,对背景区域使用限制对比度直方图均衡化增强。实验结果表明,与当前主流算法相比,本文算法主观效果更佳,且峰值信噪比、结构相似性、信息熵等客观评价参数均有所提升。We propose an infrared image enhancement algorithm based on an improved local contrast(LC)significance detection algorithm and two-area histogram equalization to improve the visual effect of an infrared image,highlight detailed information,and suppress noise.First the LC saliency detection algorithm was combined with local entropy weighting to obtain the saliency map.Then,the saliency map was adaptively segmented into foreground and background regions by the K-means algorithm.Finally,the foreground sub-histogram was equalized using a local variance-weighted distribution.The background region was enhanced using contrast-limited adaptive histogram equalization.Experimental results showed that the subjective effect of the algorithm in this study was better than the current mainstream algorithms,and the objective evaluation parameters,such as peak signal-to-noise ratio,structural similarity,and entropy,were also improved.

关 键 词：局部熵 显著性检测 K-means算法 局部方差 直方图均衡化 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]