A Survey of Multi-Focus Image Fusion Methods

As an important branch in the field of image fusion, multi-focus fusion technique can effectively solve problem optical lens depth field, making two or more partially focused images fuse into a fully image. In this paper, methods based on boundary segmentation was put forward as group method. Thus, novel classification method algorithms is proposed: transform domain methods, deep learning and combination methods. addition, subjective objective evaluation standards are listed, eight common indicators described detail. On basis lots literature, paper compares summarizes various representative At end some main limitations current research discussed, future development prospected.