Super-Resolution for Hyperspectral and Multispectral Image Fusion\n Accounting for Seasonal Spectral Variability

Abstract

Image fusion combines data from different heterogeneous sources to obtain\nmore precise information about an underlying scene. Hyperspectral-multispectral\n(HS-MS) image fusion is currently attracting great interest in remote sensing\nsince it allows the generation of high spatial resolution HS images,\ncircumventing the main limitation of this imaging modality. Existing HS-MS\nfusion algorithms, however, neglect the spectral variability often existing\nbetween images acquired at different time instants. This time difference causes\nvariations in spectral signatures of the underlying constituent materials due\nto different acquisition and seasonal conditions. This paper introduces a novel\nHS-MS image fusion strategy that combines an unmixing-based formulation with an\nexplicit parametric model for typical spectral variability between the two\nimages. Simulations with synthetic and real data show that the proposed\nstrategy leads to a significant performance improvement under spectral\nvariability and state-of-the-art performance otherwise.\n