Remote sensing image fusion based on dictionary learning and sparse representation

Abstract

It is a tough challenge to find a remote sensing image fusion method which can acquire spatial and spectral information as much as possible from panchromatic (PAN) image and multispectral (MS) image. Sparse representation (SR) can realize remote sensing image fusion better than other popular methods, which is a powerful tool for dealing with the signals of high dimensionality. In addition, to gain better fusion results without color distortion, this paper propose a remote sensing image fusion algorithm with SR and color matching in stead of the intensity hue saturation (IHS) color model and Brovey transform. The experimental results show that proposed method can make fused image with both better spatial details and spectral information compared with three well-known methods.