A MULTI-LEVEL SYNERGISTIC IMAGE DECOMPOSITION ALGORITHM FOR REMOTE SENSING IMAGE FUSION

Image fusion is a technique for improving the image quality, and image decomposition is one of a common method of image processing. In this paper, a novel multi-level synergistic image decomposition algorithm is proposed for the fusion of remote sensing images. The fusion framework decomposes different input images at different level to extract the salient and low-rank parts, respectively. The salient parts are fused using a designed weighted fusion method based on nuclear-norm, and the low-rank parts are fused using weighted average strategy. The proposed method shows a superior fusion performance in the compared experiments with PCA and GS methods.