基于色彩恒常性的彩色图像增强改进算法  

Improved Color Image Enhancement Algorithm Based on Color Constancy

作　　者：夏思宇[1] 李久贤[1] 夏良正[1] 

机构地区：[1]东南大学自动控制系,南京210096

出　　处：《南京航空航天大学学报》2006年第B07期54-57,共4页Journal of Nanjing University of Aeronautics & Astronautics

基　　金：浙江省长三角联合攻关项目(2005E60007)。

摘　　要：将色彩恒常性算法应用于彩色图像增强中,在分析McCann Retinex算法的基础上．针对迭代参教的选取问题作出了改进。改进的算法基于一种迭代截止条件,对参数进行自适应选取,避免了人工设定,并同时减少了整个算法的运算量。实验结果表明,改进后的算法对彩色图像的颜色、亮度、对比度处理的结果符合人眼视觉系统的感知特性,应用于彩色人脸检测的预处理后,人脸正确检测率提高了1．5%。The problem of color image enhancement is analyzed. The Retinex theory is applied to color image enhancement and a modification for McCann Retinex algorithm is presented. This method can automatically determine the value for the parameter, instead of looking for fixed value, and decrease the computation amount. Experimental results confirm the effectiveness of the algorithm in color image enhancement and the images obtain better rendering by human visual systems. After using the algorithm in the preprocessing for face detection in color images, the detect rate increases by 1.5%.

关 键 词：彩色图像 色彩恒常性 图像增强 RETINEX算法 

分 类 号：TN911.73[电子电信—通信与信息系统]