一种基于灰度冗余的融合的图像增强算法  

An Algorithm of Image Enhancement for Fusion Based on Gray Redundance

作　　者：乔闹生[1,2] 叶玉堂[1] 黄永林[1] 刘霖[1] 陈镇龙[1] 方亮[1] 

机构地区：[1]电子科技大学光电信息学院,成都610054 [2]湖南文理学院物理与电子科学学院,湖南常德415000

出　　处：《光子学报》2009年第5期1288-1292,共5页Acta Photonica Sinica

基　　金：国家重点实验室基金(9140C1406020708);总装基金项目(9140A01020507DZ02);广东省东莞正业产学研项目-PCB检孔机(0710);广东省东莞正业产学研项目-外观检查机(0705-1);广东省东莞正业产学研项目-Xray检查机(0705-2)资助

摘　　要：为了改善原图像直方图像素灰度分布集中于低灰度级,与经过常规直方图均衡化处理的图像直方图像素灰度分布集中于高灰度级且均具有灰度冗余的图像增强效果,提出了一种新的图像增强算法.对原始图像进行常规直方图均衡处理,并对原始图像与进行了常规直方图均衡处理的图像分别在尽可能大的范围内进行映射,使两者变换后都尽可能去除灰度冗余而有较大的动态范围.取变换后的两图像各自的优点,融合成一幅效果最好的图像.实验结果验证了理论分析的正确性.A novel image enhancement algorithm was presented in order to slove that both the pixels in histogram of the original image,which were concentrated on low gray level,and the pixels in histogram of the image after the histogram egualization, which were concentrated on high gray level, had gray redundance. The original image was processed by traditional histogram equalization. Both the original image and the image after histogram equalization were mapped to the range as far as possible, which made the image lose gray redundance and have dynamic range. And, the merit of tow images after their changes were obtained to fuse an optimum image. The results of experiment verified the correctness of the theory analysis.

关 键 词：图像增强 灰度冗余 直方图均衡 图像融合 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]