The research activity of the IPRS team focuses on the problem of the extraction of environmental information (either thematic, geophysical and/or biophysical) from remote-sensing Earth-Observation (EO) images, i.e., from images acquired by satellite or airborne sensors. The present fast development of the remote-sensing technology presents a great potential from an environmental viewpoint, thanks to the capability to provide a repetitive synoptic view of the ground area of interest. However, this huge amount of data calls for the development of analysis and processing techniques able to extract from the available images the environmental data required by the end user. The image-processing and pattern-recognition disciplines jointly constitute a natural framework for the formulation of this information-extraction problem, providing steady methodological tools to address complex image-analysis tasks.
The main applications we are involved in are: i) resource management: by "resource" we mean inland and coastal waters, agricultural lands, forestry. The goal is to assess the quantity, the quality, and the (spatial and temporal) evolution of these resources; ii) disaster prevention and monitoring: using land cover, land use, infrastructure and risk maps, prior knowledge can be acquired in order to mitigate natural disasters such as, for instance, floodings, landslides, or forest fires. In addition, after such an event occurs, remote-sensing information turns out to be very useful in order to assess the resulting damages and to study the evolution of the event itself.