A Computer Vision System for Chess Game Tracking

In this paper, we present a real-time system that allows the detection of the moves of a chess game. In the proposed approach, each captured video frame, from a RGB webcam positioned over the chessboard, is processed through the following steps; the detection of the corner points of the chessboard grids, geometric rectification, chessboard position adjustment, automatic camera exposure adjustment, intensity adjustment, move detection and chessboard drawing. All steps were implemented in MATLAB programming environment without using any chess engine. The proposed approach correctly identified 162 of 164 moves in 3 games played under different illumination conditions.