一种基于区域特性的红外与可见光图像融合算法  

Fusion Algorithm of Infrared and Visible Images Based on Region Feature

作　　者：叶传奇[1,2] 王宝树[1] 苗启广[1] 

机构地区：[1]西安电子科技大学计算机学院,西安710071 [2]河南科技大学电子信息工程学院,河南洛阳471003

出　　处：《光子学报》2009年第6期1498-1503,共6页Acta Photonica Sinica

基　　金：国家自然科学基金(60702063);总装预研项目(513060302)资助

摘　　要：提出了一种基于区域分割和àtrous小波变换的红外与可见光图像融合算法.首先,对红外与可见光图像进行区域分割及区域关联,并按关联映射图所划分区域提取红外与可见光图像的的能量信息及梯度信息;然后,对红外与可见光图像进行多尺度àtrous小波变换分解,分解后的低频部分按照文中所提出的区域能量比和区域清晰比指标进行区域融合,高频部分采用绝对值取大算子进行融合;最后进行重构得到融合图像.结果表明,该算法既可保持可见光图像的光谱信息,又可有效获取红外图像的热目标信息.An fusion algorithm of infrared and visible images based on region segmentation and à trous wavelet transform was proposed. Region segmentation and region joint were done for the infrared and visible images. Energy information and gradient information of the infrared and visible images were extracted according to the regions generated with joint map. And muhiscale decomposition of the infrared and visible images was done with the à trous wavelet transform. The low-frequency coefficients were fused with the region based rules according to the ratio of region energy（RRE） and ratio of region sharpness （RRS） presented,and the high frequency coefficients were fused with larger absolute value operator. The fused coefficients were reconstructed to obtain the fusion image. Experimental results show that the presented algorithm can hold spectrum information of the visual image, and effectively get hot object information of the infrared image.

关 键 词：图像融合 区域分割 à TROUS小波变换 区域关联 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]