基于色彩恒常性的彩色图像增强改进算法

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%。

Improved Color Image Enhancement Algorithm Based on Color Constancy

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 au- tomatically 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 en- hancement 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%.