Robot checker player with computer vision

The purpose of this project is to study and design a robot system that can play a specially modified checker game with people. With a "vision" covering the entire game board, it monitors and memorizes every move made by the opponent. Based on its pre-programmed decision making capabilities, the system can generate the appropriate countermove by utilizing its arm. Pattern recognition and image processing techniques were used for detecting each move made by the opponent. A video camera was used to obtain an image of the game board and information in the image was processed by a video processing unit so that the patterns formed by the checker pieces can be recognized. To achieve this goal, an IBM personal computer was used as the host computer. The host computer was programmed to issue commands to the video processing unit and the robot arm. It receives data from the video processing unit, manipulates the data, makes decisions on the proper countermoves, and sends commands to the robot arm. The contents of this thesis are: 1. Design and development of the system. 2. Operational characteristics of the vision subsystem including the video processing unit. 3. The algorithm of detecting moves. 4. Description and explanation of the five-axis robot, TeachMover. 5. The hardware configuration of the robot arm and its connection with the host computer. 6. The algorithm of making robot arm countermove decision. Design and development of the system has resulted in a robotic system with computer vision capable of playing a checker game with people.