SAR and Infrared Image Fusion based on Latent Low-rank Representation

Abstract

To solve the problems of image information loss and spectral distortion during the fusion of SAR and infrared images, this paper proposes a SAR and infrared image fusion method based on Latent Low-Rank Representation (LatLRR). First, the method uses Non-Subsampled Contourlet Transform (NSCT) to obtain the low-frequency and high-frequency information of the source image. Then, the low-frequency information determines the fusion weight of the low-frequency part and the high-frequency uses LatLRR to extract low-rank components for adaptive weighted fusion. Finally, uses inverse NSCT transformation on the fusion coefficients to obtain the fusion image. Compared with other typical fusion methods, the proposed method has better visual effects, and the objective evaluation parameter values are also improved.