Hyperspectral pansharpening based on modulation of pixel spectra

This paper presents a novel image fusion methods, suitable for the sharpening of a hyperspectral (HS) image by means of a panchromatic (Pan) observation. The HS bands are expanded to the scale of the Pan image and sharpened by modulating HS pixel vectors by the ratio of the Pan image to its lowpass-filtered version. The main advantage of the proposed approach is that the spectral angle of pixel vectors are preserved after fusion is accomplished. The modulating factor is damped to avoid over-enhancement with consequent loss of spatial fidelity. Experiments carried out on a Hyperion + ALI dataset demonstrate that the proposed method favorably compares with other pansharpening approaches extended to hyperspectral data.