Infrared and visible image fusion algorithm based on multi-scale transform

Abstract

For the conventional infrared and visible image fusion algorithm with poor contrast, blurred target outline, and loss of texture detail information under low illumination conditions, an infrared and visible image fusion algorithm based on multi-scale transform is proposed. Firstly, the adaptive enhancement method based on guided filtering with dynamic range compression and contrast recovery is utilized to improve the visibility of the dark region part of the visible image. Secondly, the cross-bilateral filtering multi-scale decomposition is performed on the image to be fused to obtain the base layer image and the multi-layer detail layer image. The fusion method combining the absolute value taking larger strategy and guided filtering is used to fuse the base layer images. A method based on constructing a saliency map and weight map are proposed to fuse detailed images of each layer. Finally, the fused base and detail layers are summed to obtain the final fused images. The experimental results show that the method generates fused images with clear targets and essential details and has better visual effects and fusion accuracy than other methods.