亮度和对比度调整的内窥镜图像增强技术  

Endoscopic Images Enhancement Technique with Iuminosity and Contrast Adjustment

作　　者：李凯 徐超[1] 冯博 刘鸿瑞 LI Kai;XU Chao;FENG Bo;LIU Hong-rui(School of Electronic Information Engineering,Anhui University,Hefei 230601,China)

机构地区：[1]安徽大学电子信息工程学院,合肥230601

出　　处：《小型微型计算机系统》2022年第11期2375-2380,共6页Journal of Chinese Computer Systems

基　　金：国家重点研发计划基金项目(2019YFC0117800)资助。

摘　　要：为了提高医用内窥镜图像的血管和组织对比度,本文提出一种融合技术.该方法分为两部分,第1部分基于剪裁直方图的加权分布伽玛校正(AGCWHD),利用负像策略实现亮图像增强,采用截断伽玛值的方法增强暗图像,基于对比度强弱增强中等亮度图像.之后采用改进的离散小波变换的奇异值分解(DWT-SVD)技术实现亮度增强.第2部分基于Lab颜色空间对奇异值均衡图像的L分量执行对比度受限的自适应直方图均衡(CLAHE)增强局部对比度.在实验室自建的LEI_D数据集上基于暗、中等和亮图像3类,将该方法与其他6种现有方法进行主观和客观分析,结果表明提议方法表现出良好的亮度调节和对比度增强效果,同时能够很好地保持色彩和血管及组织的边缘细节.In order to improve the contrast of blood vessels and tissues of medical endoscopic images,this paper proposes a fusion technology.The method is divided into two parts.The first part is based on the weighted distribution gamma correction of cropped histograms(AGCWHD),which uses the strategy of negative images to enhance the bright images,and the method of truncated gamma value is employed to enhance the dimmed ones,according to the contrast strength enhance medium brightness images.Then,this paper presents a modified Discrete Wavelet Transform-Singular Value Decomposition(DWT-SVD) technology to achieve luminosity Enhancement.The second part is based on the Lab color space to perform contrast limited adaptive histogram equalization(CLAHE) in the luminosity channel(L) of the singular value equalized image to enhance local contrast.Based on the three types of dimmed,medium and bright images on the LEI_D database built by the laboratory,the proposed method and other existing methods are subjectively and objectively analyzed.The results show that the proposed method exhibits good luminosity adjustment and contrast enhancement effects,while maintaining color and blood vessel edge details.

关 键 词：亮度 对比度增强 截断伽玛值 负像 自适应伽玛校正 奇异值分解 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]