Intelligent system for intercepting moving objects

Abstract

This paper presents an intelligent system for intercepting moving objects. Air hockey is an interesting table game for two players competing to score the opponent's goal. A movement interceptor that consists of vision system, prediction-decision system and Servo pneumatics system has been developed and tested to intercept a moving puck in a prototype of air hockey game. The movement of the object is tracked with a vision system which uses a low cost USB web camera. Statistical background modeling is used to identify the objects position. Object centroid from vision system is used to predict the movement trajectory. Straight line movement estimation method is used to estimate the position where the object will arrive on the other end to be intercepted. Output of prediction i.e. target position is used to drive double acting cylinder in the pneumatics system to intercept the moving object. In order to reach the target position accurately with shortest time possible, embedded PID controller was developed. Simulation and experimental results shows that the prototype built is capable of intercepting objects moving at slow to medium speeds.