ABSTRACT
The original deœnition of hyperspectral remote sensing, also known as imaging spectrometry/ spectroscopy, refers to the acquisition of images in hundreds of contiguous spectral bands to obtain a high-spectral-resolution data for each pixel of the scene [1]. In reality, this concept is much more related to the ability of the sensors to measure narrow absorption bands rather than the number of bands. For example, despite the difference in the number of bands, the Airborne Visible Infrared Imaging Spectrometer (AVIRIS) with 224 bands (400-2500 nm) and the Compact High Resolution Imaging Spectrometer (CHRIS)/PROBA with 62 bands (410-1000 nm) are both hyperspectral sensors because they have contiguous bands in each spectral range with bandwidths of approximately 10 nm. This bandwidth allows adequate measurement of most of the narrow spectral features that appear in surface component spectra.
