平移不变小波医学图像融合方法

A medical image fusion algorithm based on translation invariance wavelet transform

针对传统小波变换融合算法对细节信息的丢失问题,提出了一种新的基于平移不变小波变换的医学图像融合算法,采用灰度加权平均法进行低频部分融合;高频部分采用基于梯度能量的加权融合规则。实验结果表明,与传统的小波变换方法相比,文中方法融合效果更加理想,较多地继承了两幅源图像的重要信息,更好地描述了图像的细节部分,更具有实用性。

To solve the problem of losing details in traditional image fusi wavelet transform, a new algorithm based on translation invariance wavelet the paper. The gray weighted average method is used to determine the low the fused image while the ggradient energy weighted fusion rule is used to components. The experimental results show that the fused image can inherit the two original images and describe more details. Being compared with transform methods, it is more practical. on algorithms based on transform is proposed in -frequency component of fuse the high-frequency s the key information of the traditional wavelet