Optical remote sensing analyzes varying electromagnetic radiation (spectral properties) in the visible - near infrared, shortwave infrared and thermal infrared spectral region (from 0.4 to 14 microns), reflected from different targets on the Earth's surface. Hyperspectral remote sensing, also known as imaging spectroscopy, is based on the analysis and evaluation of the reflected (also emitted) radiation detected by a high number of narrow, contiguous and continuous spectral bands. The detailed spectral characterization of surface absorption features provided by imaging spectrometers enables to use robust inversion algorithms for the retrieval of bio- and geochemical information over the imaged area.
Research in this topic is mainly focused on developing new and automatic methods for hyperspectral imagery analyses, including spectral modeling, extraction of detailed surface information and determination of Earth surface compositional information for geosciences and environmental applications. Calibration and validation of the algorithms developed is supported by extensive field campaigns to acquire ground-truth data at all scales of applications. We work at variable spatial scales of observation, from the laboratory scale (<cm) to field measurements (~10cm-10m), to landscape scales with sensors mounted on variable airborne platforms (drones, UAVs, planes), up to regional scale for mapping and monitoring approaches based on satellite imagery. Present activities focus in test sites in North-Eastern and Southern Europe, Mediterranean areas, Southern Africa, Brazil, Australia, and China for the following application fields: