Research on Pedestrian Tracking in Complex Environments Using UAV System
摘要
In this paper, a small quadrotor UAV equipped with an RGB-D camera is selected to track the pedestrian in an unknown and complex environment. The UAV system uses the on-board sensors to sense the environment and detects the target to be tracked. The PP-PicoDet algorithm is used to solve the problem of rapid detection of target pedestrian detection, then key point detection is used to solve the problem that the target is obscured and unable to locate the target. At the same time, the EGO-Planner planner is improved to realize the autonomous obstacle avoidance of the UAV system in complex environments and the pedestrian tracking. Finally the pedestrian tracking framework is implemented and verified through simulation experiments.