Infrared and Visible Image Fusion using a Deep Learning Framework

In recent years, deep learning has become a very active research tool which\nis used in many image processing fields. In this paper, we propose an effective\nimage fusion method using a deep learning framework to generate a single image\nwhich contains all the features from infrared and visible images. First, the\nsource images are decomposed into base parts and detail content. Then the base\nparts are fused by weighted-averaging. For the detail content, we use a deep\nlearning network to extract multi-layer features. Using these features, we use\nl_1-norm and weighted-average strategy to generate several candidates of the\nfused detail content. Once we get these candidates, the max selection strategy\nis used to get final fused detail content. Finally, the fused image will be\nreconstructed by combining the fused base part and detail content. The\nexperimental results demonstrate that our proposed method achieves\nstate-of-the-art performance in both objective assessment and visual quality.\nThe Code of our fusion method is available at\nhttps://github.com/hli1221/imagefusion_deeplearning\n