A genetic approach to Pan-sharpening of multispectral images

This paper presents a novel image fusion method, suitable for sharpening of multispectral (MS) images by means of a panchromatic (Pan) observation, based on a modified generalized intensity-hue-saturation algorithm: the MS bands expanded to the finer scale of the Pan image are sharpened by adding the spatial details derived from a difference image, which is calculated by the Pan image and a linear combination of the MS bands. Since a direct, unconditioned injection of Pan details gives unsatisfactory results, a new injection model is proposed, which provides the optimum injection by maximizing a global quality index of the fused product. The optimum injection is driven by an index function capable to measure radiometric, geometric and spectral distortions in the fused images. Fusion tests are carried out both on spatially degraded data to objectively compare the proposed scheme to the most promising state-of-the-art image fusion methods (included two commercial software solutions), and on full resolution image data to visually assess the performance of the proposed genetic image fusion method.