The Autonomous Flight of an Indoor Quadrotor Using Onboard Vision

In this paper, a vision detection and tracking system is designed to force a quadrotor to achieve the automation flight in the indoor environment based on the on-board vision system. The LED lamp belt is installed on the wall inside as the tracking target for the quadrotor. The coordination control law is designed to track the LED lamp in the straight part as well as the bent part. Moreover, the converging analysis of the tracking control algorithm is given and collision avoidance between the quadrotor and the tunnel is guaranteed. Furthermore, the vision detection system is designed to provide the input signal for the controller, which is obtained by solving relative pose between the quadrotor and the LED. Finally, the proposed control and detection schemes are validated in the experiments on a real quadrotor platform.