Infrared and Visible Image Fusion via Feature-Oriented Dual-Module Complementary

With the industrial demand caused by multi-sensor image fusion, infrared and visible fusion (IVIF) technology is flourishing. In recent years, scale decomposition methods have led trend for feature extraction. Such methods, however, low time efficiency. To address this issue, paper proposes a simple yet effective IVIF approach via feature-oriented dual-module complementary. Specifically, we analyze five classical operators comprehensively construct spatial gradient capture module (SGCM) brightness supplement (IBSM). SGCM, three kinds of maps are obtained, respectively, introducing principal component analysis, saliency, proposing contrast estimation considered relative differences information covered input images. These later reconstructed through pyramidal transformation to obtain predicted image. The IBSM then proposed refine missing thermal in Among them, improve measurement applied exposure modalities, namely, grayscale images (2D gradient) well-exposedness. former responsible extracting fine details, latter meant locating regions. Experiments performed on public datasets demonstrate that method outperforms nine state-of-the-art terms subjective visual objective indicators.