Remote Sensing Image Fusion Based on Nonsubsampled Contourlet Transform and PCA

Abstract

This paper proposes a new method of fusing panchromatic and multispectral images based on NSCT and PCA. The first principle component is obtained by applying PCA on multispectral images. Then fusing the first principle component and low frequent coefficients of panchromatic images after transformed by NSCT. Substituting the fusion data performed by inverse NSCT for the first principle component and applying the inverse PCA transform. Experimental results show that the proposed method can effectively preserve spectral information and improve spatial information than traditional HSI, PCA and NSCT, and has better subjective visual effect.