ARPool (Augmented Reality Pool) is the next evolution in the game of pool. The system combines advanced computer algorithms, including computer vision, graphics, physics simulation, and even artificial intelligence, to enhance a pool player’s experience. In this way, ARPool represents the convergence of traditional games such as pool, with recent advances in computer gaming technology.
Getting into the technical details, ARPool is what is known as a “projector-camera” system. A data projector and high resolution video camera are mounted on the ceiling, directly above and pointing down towards the surface of the table. Custom computer vision algorithms have been developed to analyze the balls on the table, determining their position as well as their identity (such as cue ball, etc.). When the player lines up to take a shot, the cue position is also recognized in real-time, and the ball trajectories that will result in that angle are calculated. Finally, the data projector renders the ball trajectory information directly onto the table surface, giving the player information about the success of the shot. All this happens in realtime, so that the player can adjust his shot according to the system display.
The benefit of the system is two-fold. First, it can be used as a training aid, helping novice and even experienced players determine the best shot to take in a given scenario. The instant replay can then be used to diagnose any errors that may exist in the player’s technique. Second, the system enhances (indeed “augments”) the game of pool, giving it all of the appeal of a state-of-the-art video game.
ARPool has been in development for several years by the RCV (Robotics and Computer Vision) Lab at Queen’s University and is a spin off of an earlier project Deep Green, a robot capable of playing pool. My involvement in the project began in May 2012 when I helped prepare the system for a demonstration at OCE Discovery in Toronto. Since then I have made many improvements to the system mainly regarding the overall architecture and vision algorithm improvements.