A robust vision system for a UAV transporting cargoes between moving platforms

Motivated by the 2013 International UAV Innovation Grand Prix (UAVGP), we design and implement a real-time vision system for an unmanned helicopter to transfer cargoes between two moving platforms fully autonomously. The key algorithms in the vision system include ellipse detection, ellipse tracking and single-circle-based pose estimation. Comprehensive experiments have verified the efficiency, accuracy and robustness of the algorithms. The developed vision system has helped us achieving a great success in UAVGP. Our team was ranked first in the final round competition.