An information fusion approach for glacier mapping using multi-source remote sensing data

Semi-automatic developed techniques for mapping glaciers are essentially based on handling each source of information separately. These sources of information, formed often by satellite images and DEM (Digital Elevation Model), are acquired from different sensors at different dates and usually do not have the same spatial resolution. This aspect has been considered in this remote sensing based study to distinguish between three regions (glaciers, moraines and lakes) in a test area situated into the Alps mountains. A semi-automatic multi-source approach for the glacier detection is used which integrates the inputs from SPOT satellite images and DEM into a common space representation. Information extracted from these sources is aggregated using the belief function theory in order to take into account uncertainty and imprecision of the inputs.