Infrared and visual image fusion based on multi-scale feature decomposition

Many applications have benefited from infrared (IR) and visual (VI) image fusion technology. To integrate more useful information from the source images, in this paper, we propose a novel IR and VI image fusion method based on multi-scale feature decomposition (MSFD). The main advantage of MSFD is that it separates image features more effectively. Firstly, MSFD is introduced to decompose the source images into base layers and several detail (feature) layers. Then, the average fusion rule is applied to fuse the decomposed base layers to ensure the robustness of the proposed method, and the sum fusion rule is used to fuse the decomposed feature layers to ensure that more feature information can be transferred from the source images to the fusion result. Finally, the fusion result is reconstructed by adding the fused base layer and the fused feature layers. Experimental results demonstrate the advantages of our proposed method over other compared methods in term of visual observation and evaluation metrics.