Rapid mapping of high resolution SAR scenes

This paper describes a semi-automatic procedure for cartographic mapping using high resolution SAR and interferometric SAR data. Various two-dimensional features are extracted and combined in order to achieve a basic yet effective recognition of the elements in the scene. Many relevant elements of the landscape are automatically extracted without requiring any deep interaction with the operator. Being based on geometric models assuming regularity of shapes and patterns, the procedure is well suited for detecting man-made features, such as the road network (outside and inside human settlements) and built-up areas. It can be used, however, to extract natural features, focusing on different geometric models. Moreover, extracted elements of the scene can be grouped into higher level ones, such as crossroads, bridges and overpasses, through data fusion at the feature level, because the procedure is characterized by a multi-scale, object-based approach.