一种基于累积分布的低照度彩色图像增强算法  

A Novel Algorithm Based on Cumulative Distribution for Low Illumination Color Image Enhancement

作　　者：冯清枝[1] 虎添翼 FENG Qing-zhi HU Tian-yi(Department of Audio and Visual Material Examination Technology, Criminal Investigation Police University of China, Shenyang 110854, China Qingyang Police Station, Kunshan Police Bureau, Kunshan 215300, China)

机构地区：[1]中国刑事警察学院声像资料检验技术系,沈阳110854 [2]昆山市公安局青阳派出所,江苏昆山215300

出　　处：《光电技术应用》2017年第5期57-61,共5页Electro-Optic Technology Application

基　　金：国家自然科学基金(61603415)

摘　　要：为了解决低照度环境引起的彩色图像退化问题,提出一种新的彩色图像增强算法。首先在YUV色彩空间下采用彩色双边滤波估计照度分量;然后在提取反射分量时,构造一个与累积分布相关的对比度调节函数,自适应地增强图像的细节信息;最后采用线性彩色恢复算法恢复增强图像的色彩信息。实验结果表明,该算法能够明显地提高低照度彩色图像的整体视觉效果,细节鲜明突出,色彩真实自然。To resolve the degradation of color images caused by low illumination condition, a novel algorithm to improve the visibility of low illumination color images is proposed. At first, the illumination component is estimated using the color-bilateral filtering in YUV color space. And then, while the reflectance component being calculated, a contrast regulation function based on cumulative distribution is introduced to adaptively emphasize minute features of images. At last, a linear color restoration process is utilized to determine color values of the enhanced image. Experimental results show that the proposed algorithm is more effective and practicable, and the enhanced image can achieve an excellent global visual effect with clear details and natural colors.

关 键 词：图像增强 RETINEX理论 双边滤波 累积分布函数 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]