Tracking a single cyclist during a team changeover on a velodrome track with Python and OpenCV

Team pursuit races are contested by teams of 4 cyclists in a velodrome. The lead rider in each team experiences the highest aerodynamic drag. Team-mates alternate the lead to delay fatigue. The path taken during the changeover is critical to finishing time. An automated video tracking system was developed to track the riders during changeover. It will be used by coaches to refine the team’s performance during changeover. The system uses a single camera in burst mode, and employs background subtraction, edge detection and predictive tracking. The system is implemented in Python and OpenCV.