一种基于提升小波分解图像融合方法

Image fusion algorithm based on adaptive lifting wavelet decomposition

提出一种新的基于自适应提升小波变换的图像融合方法。基于信号局部结构特征的相关性和方向性,采用自适应提升策略构造小波;利用提升小波对图像的低频和高频分量进行分解,在两幅图像配准的前提下,将小波系数合并,进行小波逆变换,得到融合后的图像;引入融合对称度来判断融合方法的性能。比较融合后图像的熵、相关函数和融合对称度性能指标,提升小波变换方法优于直接平均法和Symm lets小波变换方法。

A novel image fusion algorithm（IFA） based on adaptive lifting wavelet transform （ALWT） was proposed. The adaptive lifting wavelet transform was mainly based on the correlation of local structure of the signal and the direction. It utilizes this information to construct adaptive wavelets decompositions via lifting scheme and reconstruction. Two registered images were decomposed by adaptive lifting wavelet transform, and then the coefficients were merged. The fusion image was gotten by coefficients reconstruction. In addition, the concept of fusion symmetry （FS） was introduced as a measure of evaluating performance of fusion algorithms. The smaller the FS the more symmetric was the fused image. Comparing the performance of entropy, correlation function and FS, the IFALWT is better than the methods of the average and the Symmlets wavelet.