Down-scaling of satellite hyperspectral images for monitoring croplands

Remote sensing has potential to provide a cost-efficient and fast tool to map soil properties across large areas. Especially, hyperspectral image can potentially discriminate between crop residues and soils as well as vegetation. Satellite hyperspectral image has very narrow spectral bands but a coarse spatial resolution to detect soil properties and vegetation in small parcels of croplands. This study focused on improving spatial resolution of the satellite hyperspectral image preserving fine-spectral resolution through the integrated down-scaling algorithms. The spatially down-scaled hyperspectral image could show better analysis results of soil properties, crop residues, and vegetation types and enhance their mapping accuracy with less loss of spectral information.