Abstract: This paper presents a lifting wavelet transform-based image fusion method. And a different fusion rule is applied in the different frequency domains of wavelet decomposition. While low frequency coefficients are chosen, the method is based on edges of image. While high frequency coefficients are chosen, the wavelet coefficient variance and the absolute value is used to fuse the wavelet coefficient. The experimental results show this fusion scheme is effective than other fusion schemes.

Key words: lifting wavelet transform, fusion rules, image fusion

摘要： 给出了一种基于提升小波变换的图像融合方法，并对小波分解的不同频率域，分别采用不同的融合规则。选择低频系数时，是基于边缘的方法，选择高频系数时，把小波系数的方差与绝对值综合起来考虑来决定融合小波系数。实验结果表明，提出的方法融合效果要优于一般融合方法。

关键词: 提升小波变换, 融合规则, 图像融合