ABSTRACT
ABSTRACT: Airborne laser scanning has become an accepted technique for acquiring 3D geodata of the earth’s surface. The extraction of the off-terrain objects is one of the challenging topics. In this paper the off-terrain regions are separated from the terrain regions by means of geodesic image reconstruction. The key of the segmentation process is to analyze a generated sequence of morphologically filtered data to extract ground points with high probability and separate them from non-ground points. Aggregation to regions and the extraction of region properties provide the basis for 3D object extraction. Further analysis focuses the feature description for the 3D regions that provides the input for classifying and separating 3D objects, in particular buildings and vegetation regions, from the ground surface regions. The process will be continued and focused on the building outlines with vectorization and approximation. An algorithm based on a genetic algorithm is proposed to simplify and generalize the building boundaries. Building boundaries are modeled generically based on rectangular closed polygons. The aim of the approximation is to find the best rectilinear polygons which fit the building outlines. This procedure is evaluated with a data set, which was recorded by the TopScan laser scanning system with the density of about 1.7 points per square meter.
