Infrared and Visible Image Fusion Based on Oversampled Graph Filter Banks.

The near infrared (NIR), long-wave infrared (LWIR) and visible image fusion combines the interesting information of NIR and LWIR images and the details of visible images, which can provide an informative image for human visual perception and subsequent processing. In this paper, we propose a novel image fusion method based on oversampled graph filter banks. And then, the fusion rules for low frequency subbands and high frequency subbands are constructed by a new method of autoselect weighted average or maximum absolute value independently. Finally, the fusion image is obtained by applying the inverse transform on the fusion subband. The experimental results show that the proposed method outperforms the recently proposed 4 methods in terms of visual effects and performance indicators.