Multi-focus Image Fusion Based on Fuzzy and Wavelet Transform

摘要

In this paper, we proposed a new method for spatially registered multi-focus images fusion. Image fusion based on wavelet transform is the most commonly fusion method, which fuses the source images information in wavelet domain according to some fusion rules. There are some disadvantages in Discrete Wavelet Transform, such as shift variance and poor directionality. Also, because of the uncertainty about the source images contributions to the fused image, designing a good fusion rule to integrate as much information as possible into the fused image becomes one of the most important problem. In order to solve these problems, we proposed a fusion method based on double-density dual-tree discrete wavelet transform, which is approximately shift invariant and has more sub-bands per scale for finer frequency decomposition, and fuzzy inference system for fusing wavelet coefficients. This new method provides improved subjective and objectives results compared to the previous wavelet fusion methods.