Stabilization of an unmanned aerial vehicle using real-time embedded motion estimation

This paper presents an image-based velocity control method for a VTOL UAV. For increased autonomy, a realtime embedded global motion estimation process using efficient feature tracking was developed. The VTOL vehicle is assumed to be equipped with a minimum sensor suite — i.e. a camera, altitude sensor and IMU — and to hover over a featured flat plane. This paper first presents a fast feature-based global motion estimation method, and then proposes a control strategy based on this method. Finally, experimental results using a quadrotor demonstrate the performance of the proposed control strategy and feature-based motion estimation.