DMDN: Degradation model-based deep network for multi-focus image fusion
作者：
Highlights：
• The degradation model of multi-focus image fusion (MFIF) guides the construction of the network.
• The end-to-end network directly fuses the defocused images into an all-in-focus image.
• Random concatenate input order during training improves the generalization ability of the network.
• We publish an MFIF dataset containing 70 pairs of high-resolution multi-focus images.
• We report state-of-the-art fusion performance on three MFIF datasets.
摘要
•The degradation model of multi-focus image fusion (MFIF) guides the construction of the network.•The end-to-end network directly fuses the defocused images into an all-in-focus image.•Random concatenate input order during training improves the generalization ability of the network.•We publish an MFIF dataset containing 70 pairs of high-resolution multi-focus images.•We report state-of-the-art fusion performance on three MFIF datasets.
论文关键词：Image fusion,Multi-focus,Deep learning,Degradation model
论文评审过程：Received 7 December 2020, Revised 29 September 2021, Accepted 31 October 2021, Available online 24 November 2021, Version of Record 29 November 2021.
论文官网地址：https://doi.org/10.1016/j.image.2021.116554
