可避免彩色失真的图像对比度增强  

Image Contrast Enhancement for Avoiding Color Distortion

作　　者：李冠章[1] 罗武胜[1] 李沛[1] 

机构地区：[1]国防科学技术大学机电工程与自动化学院,湖南长沙410073

出　　处：《计算机仿真》2008年第10期228-231,共4页Computer Simulation

摘　　要：为了将灰度变换技术推广到RGB彩色图像增强,并克服其容易引起彩色失真的缺陷,提出了一种通用的自适应彩色图像对比度增强算法。首先采用小波变换,直接在RGB空间分离出原始彩色图像和灰度变换图像的低频部分和高频部分,然后提取原始图像的彩色信息和灰度变换图像的对比度信息,最后进行归一化处理再重构输出彩色图像。实验结果表明,算法能在不引起明显彩色失真的前提下改善图像视觉效果,与基于Retinex理论的多尺度彩色复原(MSRCR)算法相比,逼真度和运算复杂度等指标都更加优异。This paper presents a novel general method of adaptive color image contrast enhancement, which not only extends gray level transform to RGB image enhancement, but also overcomes the limitation of color distortion. Wavelet transform is adopted to separate directly low frequency and high frequency parts from original color image and gray level transform image in RGB color space. Then color and contrast information of original image and gray level transform image are picked up, finally output color image is normalized and reconstructed. Experiment results show that this algorithm improves image visual effect without introducing color distortion. Fidelity and operation complexity of this algorithm are better than that of Multi - scale Retinex Color Restoration （MSRCR） that is based on Retinex theory.

关 键 词：彩色失真 灰度变换 小波变换 多尺度彩色复原 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]