Multispectral image fusion based on the Multiwavelet and IHS transforms

Abstract

This paper introduces a fusion method to merge the IKONOS low resolution multispectral image (MS) with the IKONOS high resolution panchromatic (PAN) image based on the Multiwavelet transform. Different fusion rules are used with the Multiwavelet method to improve the fusion quality based on pixel level fusion or feature level fusion. The best three old techniques for image fusion (IHS, Brovey, and Wavelet) are also tested in this paper. The performances of the existing and the proposed methods are calculated using Correlation Coefficient and Root Mean Square Error. Here we used our method to merge the panchromatic image of IKONOS sensor (1m resolution) with its multispectral image (4m resolution). Multiwavelet based image fusion method provides richer information in both spatial and spectral domains than the Wavelet based method, and the results clearly demonstrate the advantages of this approach.