Geometric primitive extraction for 3D reconstruction of urban areas from tomographic SAR data

Abstract

In this paper we propose a method that performs an automatic extraction of geometric primitives from tomographic SAR data. This technique allows to identify surfaces by assuming they can be locally described by a piecewise planar model. It is therefore suitable to reconstruct man-made structures in urban areas. Our approach takes advantage of the spatial connectivity of points from tomographic height maps in order to retrieve planar patches from noise corrupted point clouds. We demonstrate how our method outperforms the well-known RANSAC algorithm over synthetic and experimental data.