改进成像模型的单尺度Retinex彩色图像增强算法  

Single-scale Retinex color image enhancement algorithm based on improved imaging model

作　　者：王彦林 张进 WANG Yan-lin;ZHANG Jin(College of Information Engineering,Wuhan Business University,Wuhan 430056,China;School of Computer Science and Technology,Nanjing Technology University,Nanjing 211000,China)

机构地区：[1]武汉商学院信息工程学院,湖北武汉430056 [2]南京工业大学计算机科学与技术学院,江苏南京211000

出　　处：《计算机工程与设计》2018年第11期3511-3515,共5页Computer Engineering and Design

基　　金：湖北省科技支撑计划基金项目(2013BAA104);江苏省自然科学基金项目(BK20141113);江苏省科技支撑计划基金项目(BE2012096)

摘　　要：为提高彩色图像增强的对比度,消除光晕伪影,提出一种基于改进成像模型的单尺度Retinex彩色图像增强方法。设计一个由全局照度、局部照度和反射率乘积构成的改进成像模型,将输入RGB彩色图像转换为HSV彩色图像;分别对V分量3个估计分量进行估计和伽马校正,图像H和S分量保持不变,采用改进成像模型获得增强的输出V分量;结合H、S和增强后V分量获得输出图像。实验结果表明,相比当前彩色图像增强方法,所提方法具有更好的增强性能,其输出彩色图像具有较高的全局和局部对比度,无颜色变化,有效消除了光晕伪影。To improve the contrast of color image enhancement and eliminate halo artifacts,a single scale Retinex color image enhancement method based on improved imaging model was proposed.An improved imaging model composed of global illuminance,local illuminance and reflectance product was designed,and the input RGB color image was converted into HSV color image.The three estimated components of the V component were estimated and gamma corrected.The image H and S components remained unchanged,and the enhanced output V component was obtained using the improved imaging model.The output image was obtained by combining H,S and the enhanced V component.Experimental results show that compared with the traditional color image enhancement method,the proposed method has better performance,and its output color image has high global and local contrast with no color change,effectively eliminates the halo artifacts.

关 键 词：彩色图像增强 HSV彩色空间 V分量 伽马校正 RETINEX 低通滤波器 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]