基于小波变换的一种改进融合规则的图像融合  

Image fusion based on improved fusion rules of wavelet transform

作　　者：朱健[1] 石跃祥[1] 

机构地区：[1]湘潭大学信息工程学院,湖南湘潭411105

出　　处：《计算机工程与应用》2011年第11期185-186,共2页Computer Engineering and Applications

基　　金：湖南省自然科学基金No.07JJ6115;湘潭大学博士科研启动项目(No.06QDZ23)~~

摘　　要：针对多聚焦图像,通过小波变换和基于窗口的系数绝对值和选大融合规则,存在着对噪声敏感和融合信息不完整等缺点。为了克服这些缺点,使融合后的图像包含尽可能多的源图信息,提出了一种在融合时两幅图像的系数都选取的融合规则。通过对多聚焦图像融合的实验结果比较表明,提出的融合规则的融合效果更优,可避免信息损失。Using the wavelet transform and the fusion rule based on windows for choosing the largest coefficient absolute sum,the multi-focus images will be susceptible to noise and have incomplete fusion information.This paper presents a kind of fusion rule which will select the coefficients of two images for the fusion to make the fusion images contain the information of original images as much as possible.Comparison of fusion results of multi-focus image is given.The proposed fusion rules can achieve better fusion performance than other usual fusion rules,avoiding losing information.

关 键 词：小波变换 融合规则 图像融合 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]