Infrared and visible image fusion algorithm based on interactive fusion network

In autonomous driving technology, infrared and visible light image fusion is essential to fully capture information about the complex environment around the vehicle. However, the current methods often face the challenge of inadequate fusion of complementary information and poor global interaction. To solve these problems, an image fusion method named CATfusion is proposed in this paper. This method aims to achieve comprehensive fusion and global interaction of information to optimize the fusion effect of infrared and visible images. First, CATfusion uses a hybrid network of convolutional layers and lightweight Transformer to extract features from input images, capturing local details and global context information in the images. Secondly, in order to fully integrate the global and local information of the image, an interactive fusion network is designed for the accurate fusion of the global and local features of the image, so as to improve the information richness and representativeness of the fused image. Then, the fusion features are reconstructed through the serial feature reconstruction network to generate a high quality fusion image with rich information and clear details. Finally, experiments on MSRS data sets show that the fusion effect of CATfusion is significantly improved compared with traditional feature fusion networks.