一种多色彩空间信息融合的图像增强算法  

A Multi-Color Space Information Fusion Algorithm for Image Enhancement

作　　者：党领茹 朱丹[1,2,3] 佟新鑫[1,3] 王玉良[1,3] 

机构地区：[1]中国科学院沈阳自动化研究所,辽宁沈阳110016 [2]中国科学院大学,北京100049 [3]中国科学院光电信息处理重点实验室,辽宁沈阳110016

出　　处：《微电子学与计算机》2014年第12期84-87,92,共5页Microelectronics & Computer

摘　　要：在彩色图像获取过程中,由于光照不均等原因导致采集的图像动态范围大,细节信息模糊.为此,提出了基于多色彩空间信息融合和同态滤波的彩色图像增强算法.首先,在RGB色彩空间上,分别对R、G、B分量进行同态滤波和自适应截取拉伸,消除光照影响,有效增强局部细节信息.然后,在HSV色彩空间上,利用由构造的非线性变换函数处理后的V分量,进行光照补偿.并且利用原图像H和S分量,恢复图像色彩信息.实验分析表明,本算法能够很好的压缩图像动态范围,增强亮暗区域的细节信息,并保持颜色的一致性.In the process of color image acquisition,due to inhomogeneous illumination and other reasons,the collected images have features of large dynamic range and details blur.So this paper proposes a method for enhancing the color images based on information fusion of multi-color space and homomorphic filters.First,in RGB color space,the three components are processed by using the homomorphic filtering and the adaptive clipping stretch methods respectively to eliminate illumination effect and enhance the local details effectively.Then,in HSV color space,the V component of modified by the nonlinear transformation function structured is used to make illumination compensation to the image.And the H and S component of the original image is used to restore the color information.The experimental results suggest that the proposed method shows better in the aspects of dynamic range compression,enhancement of the detail of bright and dark areas and keeping the color consistency.

关 键 词：图像增强 同态滤波 色彩空间 光照补偿 

分 类 号：TN391[电子电信—物理电子学]