Change detection for road condition mapping based on dry and rainy season VHR dual-pol radar imagery over tropical areas. Application to Chad

Abstract : New Very High Resolution dual-polarization radar sensors in X-band (TerraSAR-X and Cosmo-Skymed) are offering technical opportunities for the detection of road condition changes between dry and rainy period in tropical areas. We applied classification methods (fuzzy logic and learning algorithms), multi-incidence and optical-radar combination in order to estimate surface elements (relief, flooded areas), soil properties (wetness, surface roughness) and road conditions from multi-date acquisitions over SW Chad. These estimations are calibrated with data measurements on the field. They are included in a road condition service designed for military, humanitarian and industrial logistics.