Multi-focus image fusion using multi-structure top-hat transform and image variance

Image fusion is a process of combining complementary information from multiple images of the same scene into an image, so that the resultant image contains a more accurate description of the scene than any of the individual source images. In this paper, an algorithm for multi-focus image fusion based on multi-structure top hat operator and image variance is proposed. The fusion process contains the following three steps. First, the multi-focus images are fused using multi-structure top hat contrast operator. Then, the intermediate fused images are segmented into smaller blocks. Finally, the fused image is constructed by comparing the segmented blocks of the intermediate images using image variance. The performance of the method has been extensively tested on several pairs of multi-focus images and compared quantitatively with existing methods.