Quality evaluation of pansharpening techniques on different land cover types

Nowadays, high spatial resolution remote sensing systems collect simultaneously a low resolution multispectral image plus a high resolution panchromatic image. The objective of pixel level fusion methods is to combine this information in order to obtain a new multispectral image that exhibits the spectral characteristics of the multispectral image and the spatial resolution of the panchromatic image. There are many image pansharpening methods that have been proposed in the literature however, in general, none of them exhibits the best performance for all type of images. Thus, in this paper we present a comparative analysis of different fusion methods applied over a database of Geoeye-1 images having different types of land covers. We have analyzed qualitatively and quantitatively the pansharpened images in order to select those methods providing the best performance to each particular type of cover.