Infrared and visible image fusion via detail preserving adversarial learning.

•We propose a new end-to-end model based on GANs for IR/VIS fusion.•A detail loss is designed to improve the quality of detail information under GANs.•A target edge-enhancement loss is designed to sharpen the edge of IR targets.•It can keep the thermal radiation and texture details in source images simultaneously.•Results look like enhanced IR images with clear targets and abundant textures.