The project is a realization of an automated smart carrom-playing robot that continuously analyses the board configuration using a vision feed and image processing and plays a series of best possible shots to complete a game. This leverages the knowledge of reinforcement learning in implementing strategic gameplay.
The device is capable of mimicking the human experience of playing carrom. A camera mounted on top of the carrom board feeds images to the device. The images are then mapped to a two-dimensional coordinate system to identify the centers of carrom pieces and holes. This information is used to train a machine-learning algorithm to come up with the best play. The algorithm automatically determines the disk position, the speed, and the angle to hit an identified chess piece based on the game situation.