Meshing and Simplication of High Resolution Urban Surface Data for UAV Path Planning

Abstract

This work presents an approach to utilize high resolution surface data as a-priori information for three dimensional path planning at very low altitude. The major challenge is preserve features while reducing the amount of data to a minimum. Non significant height points are eliminated by a neighbor search, performed within a data structure generated from pseudo 3D Delaunay meshing. A comparison to an alternatively implemented simplification shows which inherent building features can be preserved. For highlighting the feasibility of the approach, the processing results of real urban surface data from the inner city of Berlin is presented and used for sampling based path planning of an unmanned helicopter.