IKONOS image fusion by minimisation of spectral distortion using MAP estimator

Abstract

Though intensity-hue-saturation (IHS) approaches for image fusion are very fast and easy to implement, they have the intrinsic problem of spectral distortion. To reduce colour distortion in IHS methods, a maximum a posteriori estimator is proposed that minimises colour distance between estimated intensity image and panchromatic image, which preserves spectral information and sharpens the reconstructed images. The simulation demonstrates that the proposed method reconstructs clearer colour images and yields good results, both subjectively and quantitatively.