区域图像融合算法在红外图像分析中的应用  

Region-based Image Fusion Algorithm in Infrared Image Analysis

作　　者：张勇[1] 陈大建[1] 

机构地区：[1]光电信息控制和安全技术重点实验室,河北燕郊065201

出　　处：《光电技术应用》2011年第3期17-20,共4页Electro-Optic Technology Application

基　　金：国家部委基金(9140C150105100C1502)

摘　　要：采用区域分裂合并算法获得待融合图像的共同区域,同时,对两幅待融合图像进行多尺度分解,根据区域相似度和融合量测指标对分解后的低频系数和高频系数进行加权平均和系数选取,经小波逆变换后得到融合后的图像。试验结果表明,基于区域的融合算法能够保留更多的细节信息。The common area of the candidate images is obtained by using the regional split and merge algorithm, meanwhile two candidate images are processed by the multi-scale decomposition, the weighted average and coefficient selection are done for the low frequency coefficients of decomposition and high frequency coefficients according to the fusion measuring indexes and the regional similarity, the fusion image is obtained by inverse wavelet transform. The results show that region-based fusion algorithm can preserve more details.

关 键 词：图像融合 小波变换 分裂合并 区域 

分 类 号：TN911.73[电子电信—通信与信息系统] TP722.5[电子电信—信息与通信工程]