Integrated navigation system with monocular vision and LIDAR for indoor UAVs

摘要

Without the Global Positioning System (GPS), indoor Unmanned Aerial Vehicle (UAV) navigation seems rather insurmountable. And most solutions call for external ground stations or otherwise beacons. But in this paper, for the problem, a set of indoor navigation system that rely solely on onboard equipment is proposed. The core of presented method is based on the evaluation system of the speed and location with monocular vision and the obstacle avoiding system with a laser radar. Compared with a variety of image processing algorithms, eventually two most suitable algorithms for optical flow and visual odometry are put forward. Experiments show that our navigation system is able to achieve stable navigation with high precision and autonomous obstacle avoidance.