Automatic image registration and color merging for SPOT5 imagery

Abstract

An automatic process for registering the panchromatic and multispectral SPOT5 imagery is presented. The registration process make used of edges and corner information to select informative image patches in the PAN image. A hierarchical local matching method is then used to find the best matched image patches in the XS image. The coordinates of the matched image patches are fitted to an affine transformation model. The merged high resolution color imagery is obtained by multiplying each spectral bands with a sharpening factor computed from the intensity values of the panchromatic image and the corresponding pixel values of the multispectral images.