基于线性变换的RGB图像色彩矫正算法

Color Correction Algorithm Based on Linear Transform RGB Image

颜色矫正是对颜色失真图像进行颜色恢复处理的必须过程。为解决图像偏色和颜色失真问题,本文提出了一种基于线性变换的图像色彩矫正的算法,方法简单易行,方便实时处理。算法总思路是调整偏色图像各颜色分量的分布使其向正常图像的分布靠拢。具体做法是,对于无参照图像的偏色图像,利用置信区间对各颜色分量进行线性调整,对于有参照图像的偏色图像,将偏色图像各颜色分量均值进行补偿使其接近参照图像的各颜色分量均值。从实验结果和Lab色域评价看,本文算法较好地实现了色彩矫正。

Color correction is a necessary process for color recovery of image color distortion.In order to solve the problem of image color cast and color distortion,this paper proposes an image color correction algorithm based on linear transformation.The method is simple and convenient,and it is convenient for real-time processing.The general idea of the algorithm is to adjust the distribution of each color component of the color cast image to the similar distribution of the normal image.For the color cast image without referential image,each color component is linearly adjusted by using the confidence interval,and for the color cast image having referential image,the mean value of each color component of the color cast image is compensated to be close to counterpart of the referential images.Through the post-correction results and the Lab color space evaluation,the algorithm achieves better color correction effect.