DSFusion: Infrared and visible image fusion method combining detail and scene information

Kuizhuang Liu, Min Li, Cheng Chen, Chengwei Rao, Enguang Zuo, Yunling Wang, Ziwei Yan, Bo Wang, Chen Chen, Xiaoyi Lv

Abstract: Highlights•The model can take into account both detail and scene information.•A novel local attention module is proposed.•We have improved the channel attention module.•Two novel loss functions are proposed.•Our model has good generalization ability.