基于小波变换的多聚焦图像融合算法  

Multi-focus image fusion algorithm based on wavelet transform

作　　者：赵立强[1] 杨大志[1] 周艳红[1] 向洁[1] 

机构地区：[1]河北科技师范学院数学与信息科技学院,河北秦皇岛066004

出　　处：《计算机工程与应用》2015年第23期184-188,共5页Computer Engineering and Applications

基　　金：河北科技师范学院博士基金(No.2011YB-006)

摘　　要：针对基于小波变换的多聚焦图像融合算法,改进融合规则和融合算子,低频分量采用以相关系数作为阈值的加权平均算法,高频分量采用基于区域特征的融合算法,并对最佳分解层数与最佳小波基的选取进行优化验证。通过对实验结果的分析,选用bior4.4小波,进行最佳分解层数小波分解,并应用改进的融合规则,在融合多聚焦图像的效果上,与其他多种融合算法相比,各项评价指标都比较理想。In view of multi-focus image fusion algorithm, fusion rules and fusion operators are improved. The low-frequency component uses the weighted average algorithm with correlation coefficient as threshold, the high-frequency component uses fusion algorithm based on regional characteristics. In addition, the optimal decomposition level and the selection of the best wavelet basis are verified. The evaluation indexes of multi-focus image fusion based on improved fusion rules and bior4.4 wavelet with optimal decomposition level, through the analysis of experimental results, are better than many other fusion algorithm.

关 键 词：多聚焦图像融合 小波变换 小波基 融合规则 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]