Fusion Image
Image fusion integrates information from multiple image sources to create a composite image exceeding the quality or information content of any single input. Current research emphasizes developing advanced deep learning models, including generative models, transformer networks, and graph neural networks, to improve fusion quality and efficiency across diverse applications such as infrared-visible image fusion, medical imaging, and autonomous driving. These advancements aim to address challenges like handling varying illumination conditions, preserving fine details, and achieving real-time processing speeds, ultimately improving diagnostic capabilities, scene understanding, and robotic perception.
Papers
IAIFNet: An Illumination-Aware Infrared and Visible Image Fusion Network
Qiao Yang, Yu Zhang, Zijing Zhao, Jian Zhang, Shunli Zhang
SSPFusion: A Semantic Structure-Preserving Approach for Infrared and Visible Image Fusion
Qiao Yang, Yu Zhang, Jian Zhang, Zijing Zhao, Shunli Zhang, Jinqiao Wang, Junzhe Chen