In this paper, we propose a full-stack system for perception-based accurate autonomous vehicle following that does not rely on vehicle-to-vehicle communication, map data, or accurate global localization. Our system consists of simple components for localization, object tracking, behaviour and waypoint management, trajectory planning, and control. We assessed the system's performance through different practical experiments using a real autonomous car. The experiments were conducted in various scenarios, including driving on- and off-road with speeds of up to 20 m/s and in the presence of dynamic obstacles.
Perception-Based Accurate Autonomous Vehicle Following in GNSS-Denied Environments
24.09.2024
5009763 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
Autonomous Flight in Unknown GNSS-denied Environments for Disaster Examination
ArXiv | 2021
|