Abstract: In order to obtain good fusion results, a novel image fusion algorithm based on wavelet similarity and weight is proposed in this paper. Firstly, wavelet similarity is used to reflect the detail structure information in source images. Secondly, contrast sensitivity function is adopted to weight the wavelet coefficient on different scales, and mages are fused by weighted similarity. Finally, the simulation experiment is use to test the performance of the algorithm. The experimental results show that the proposed algorithm can make the image fusion effect more naturally and obtain better result than traditional similarity method.

Key words: image fusion, wavelet decomposition, weighted similarity, contrast sensitivity function

摘要： 为了获得较好的图像融合效果，针对传统相似度算法存在的不足，提出一种基于小波相似度加权的图像融合算法。采用小波相似度描述图像细节结构信息，并使用对比度敏感函数对不同尺度的小波系数进行加权，然后利用加权相似度对图像进行融合，最后采用仿真实验对算法性能进行测试。实验结果表明，该算法能够进一步衡量图像间的结构相似度，图像融合过渡效果更加自然，取得了优于传统相似度方法的融合效果。

关键词: 图像融合, 小波分解, 对比度敏感函数, 相似度加权