Abstract: A new lifting 2-D wavelet transform method was presented and applied into image fusion.This lifting transform was based on the correlation of local structure of the image and direction.Two registered images were decomposed by this new lifting wavelet transform，and then merged the coefficients.The fusion image was gained by reconstruction.Comparing the performance of entropy，correlation coefficient and image definition，this lifting transform is better than the traditional wavelet transform and the tensor product wavelet lifting method in terms of the adaption，the speed of image fusion and the spatial quality.

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

摘要： 提出了一种新的二维小波提升方法，并把它应用于图像融合中。该方法利用图像局部结构的相关性和方向性，将两个配准后的图像经过提升小波分解，合并相关系数，再通过逆变换，得到融合后图像。并引入信息熵、相关系数和清晰度等性能指标对融合后的图像进行分析。实验结果表明，此提升方法在融合速度以及融合后的图像质量上优于传统小波变换和张量积小波提升方法。

关键词: 图像融合, 提升小波变换, 离散小波变换