基于改进直方图均衡化和SSR算法的灰度图像增强研究α~←  

Gray image enhancement based on improved histogram equalization and SSR algorithm

作　　者：胡倍倍 吕浩杰[1] 

机构地区：[1]河南大学物理与电子学院,河南开封475004

出　　处：《量子电子学报》2017年第3期286-292,共7页Chinese Journal of Quantum Electronics

基　　金：河南省教育厅自然科学研究项目;13A535073;河南大学科研基金项目;2012YBZR037~~

摘　　要：为了提高偏亮和偏暗图像的增强效果,提出了一种改进的直方图均衡化和SSR算法相结合的灰度图像增强方法。对原始图像进行改进直方图均衡化增强,提取低频分量进行直方图均衡化,与高频分量融合;对原始图像进行SSR算法增强;将得到的2个图像进行加权融合。结果表明:新方法得到的图像比原始图像细节特征突出,视觉效果明显改善;对比度和信息熵得到提高、亮度得到调整。在医学X射线图像上验证了其可行性,与其他增强算法相比,新方法更有利于图像的准确分析。In order to improve the enhancement effect of partial brighter and darker images, an improved gray image enhancement method is proposed, which is based on histogram equalization and single-scale Retinex （SSR） algorithm. Improved histogram equalization enhancement for original images is carried out, and the low frequency components of the original images are enhanced by histogram equalization and fused with the high frequency components. SSR algorithm enhancement is carried out to original images. Weighted fusion of the two images is carried out. Results show that the image obtained by the new method is more prominent to highlight details of the original image, and the visual effect is improved significantly. Contrast and entropy are enhanced, and the brightness is adjusted. The feasibility on medical X-ray images is verified. Compared with other enhancement algorithms, the new method is more conducive to the accurate analysis of images.

关 键 词：图像处理 图像增强 直方图均衡化 SSR算法 图像融合 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]