Kinect Board Game Tracking

Supervisor: Martin Kruliš
Intended Scope: Bachelor thesis, Master thesis, seminar work, or software project
Required Skills: C#, kinect, image analysis


One or a small set of board games which use easily distinguishable board and tokens will be selected (e.g., chess). The objective is to create a tracker that records player moves -- e.g., creates movement record in algebraic chess notation. Simple visual tracking may be difficult, so Kinect will be used instead of regular camera. Kinect provides not only visual image, but also computes depth map form IR sensor, so it makes it much simpler to detect the pieces on the board.

The main focus is on the development of the algorithms for piece tracking (given the rules of the game). However, based on the intended scope of the project, the work can be extended in several ways:

  • Game visualization - showing the state of the game being recorded on the monitor (and possibly replay the game).
  • Movement verification and undo - notify the players that the last move was not correct and allow them to fix it.
  • Comments and suggestions - provide feedback about the quality of the latest moves and offer suggestions for the next move.
  • Autocalibration - boards and player tokens may come in different sizes and shapes even for the same game (e.g., there are many types of chess pieces and boards). The system should be able to calibrate on initial game position when the pieces are at their initial locations.
  • Automatic piece movement using either magnets under the board or something even more elaborate like Arduino-controlled robotic arm, so the player can play against computer using traditional game board instead of UI in computer.
The content is available under Creative Commons BY-NC 3.0 License