采用小波分析的图像融合方法评述

Overview of Wavelet-Analysis-Based Image Fusion

图像融合是多传感器信息融合在图像处理领域的一个重要应用,以小波分析为工具是这一领域研究方法上的重大突破。在引入图像的小波多分辨分析的基础上,介绍了图像的快速小波分解,提升小波算法;小波图像融合步骤及融合过程中融合规则的选取,融合效果的客观定量评价。最后,通过对小波滤波器组选择及有代表性的融合规则的大量实验,并用熵,交叉熵,交互信息量进行定量分析和比较,给出了结论,并指出该领域今后发展的热点。

Image fusion is an important application of multi-sensor information fusion in image processing area,and using wavelet transform is a bigness breakthrough of study method in this area.Based on introducing image multi-resolution analysis,the image fast wavelet decomposition and lifting scheme are presented.This paper pays attention to the process of wavelet image fusion,the existing fusion rules and the evaluation of fusion performance.We have done a lot of experiments on the different wavelet filters and the representative fusion schemes,and use entropy,cross entropy,mutual information to evaluate the performance of them.Finally,we can draw the conclusion and point out the hotspot of this area developing in the future.