Vision Based Tracker for Dart Catching Robot
Magnus Linderoth · Lund University Publications Student Papers (Lund University) · 2008
The objective of this thesis has been to develop the foundation for a robot system that catches darts. A dart board is to be mounted on a robot. When a dart is thrown at the board, it is detected by cameras, and an algorithm predicts where the dart will hit the board. The goal is then to move the board in such a way that the dart hits at the desired coordinate, typically the bull's eye. This report describes the different components developed to realize this system, including image acquisition, camera calibration, image analysis and modeling and estimation of the dart trajectory.