改进单尺度Retinex的彩色图像增强算法  

Color image enhancement algorithm based on improved single scale Retinex

作　　者：赵冬梅[1] 李斯娜[1] ZHAO Dongmei;LI Sina(Baoshan University,Baoshan 678000,China)

机构地区：[1]保山学院,云南保山678000

出　　处：《现代电子技术》2018年第17期71-75,共5页Modern Electronics Technique

摘　　要：为了解决传统单尺度Retinex算法全部像素都采用同一空间滤波模块实施操作,容易产生彩色图像光晕以及色彩流失,彩色图像增强效果差的问题,提出改进单尺度Retinex的彩色图像增强算法。改进算法融合粒子群优化算法以及Retinex算法,基于熵值大小以及变换成HIS图像的H值大小实施匹配,采用合理的滤波模板实施运算,完成彩色图像的增强处理。通过改进SSR算法解决图像边缘亮度波动时,图像出现的光晕以及光阴影问题,提升彩色图像增强效果。实验结果表明,所提算法对室内室外图像的亮度、对比度、饱和度等内容具有较高的增强效果,能提升总体图像的质量。The same spatial filtering module is adopted for all the pixels in the traditional single scale Retinex algorithm,which is easy to produce the color reduction and halo of color images,and has poor effect of color image enhancement.Therefore,the color image enhancement algorithm based on single scale Retinex is proposed,in which the particle swarm optimization algorithm and Retinex algorithm are combined,the color images are matched on the basis of entropy value and H value transformed into HIS image,and the reasonable filtering template is adopted for calculation to enhance the color images.The SSR algorithm is improved to solve the halo and light shadow appearing in images while the luminance of image edge is fluctuating to improve the enhancement effect of color image.The experimental results show that the proposed algorithm has high enhancement effect for indoor and outdoor images in the aspects of brightness,contrast and saturation,and can improve the quality of the overall image.

关 键 词：改进单尺度 RETINEX算法 粒子群优化 SSR算法 彩色图像增强 滤波模板 

分 类 号：TN911.73-34[电子电信—通信与信息系统] TP391.41[电子电信—信息与通信工程]