Indoor Position Control of a Quadrotor UAV with Monocular Vision Feedback

This paper aims at designing an indoor position control system for a quadrotor unmanned aerial vehicle (UAV), which cruises inside a room without GPS signals. A monocular camera equipped on the quadrotor is used to observe ArUco markers that are deployed on the floor of the room, and estimate the pose of the quadrotor. The Kalman filter is applied to further improve the accuracy of pose estimation. A PID controller is designed to realize closed-loop position control. We setup a quadrotor platform to verify the effectiveness of the indoor position control algorithm. Experiments show that the monocular vision feedback module is able to provide accurate and real-time pose estimation, and the closed-loop position control algorithm successfully guides the quadrotor to track a predefined trajectory.