UAV 3-D Path Planning Based on High-Resolution DSM, DTM, and True Orthomosaic

Abstract

In this letter, it is proposed an offline, shortest, and safer path planning algorithm for a multirotor Unoccupied Aerial Vehicle (UAV). The aim is to generate a UAV trajectory that avoids specific obstacles on the ground while maintaining a safe and legal flight height. A free-collision path is generated by simulating the UAV trajectory over a restriction map, obtained contemplating buildings footprint and surroundings over a true-orthomosaic, using the A* algorithm for 2D path planning, combined with both the Digital Terrain Model (DTM) and Digital Surface Model (DSM) to analyze the altimetric safety trajectory. The sub-optimal path is smoothed by cartographic generalization, adopting the parameter of generalization equal to the binary image resolution that corresponds to the restriction map. The experiments show that the algorithm can generate a 3D path that is efficient and safe, for both the UAV and the objects on and above ground, according to the regulations to access the Brazilian airspace.