Fast outcome prediction based on slow cause estimation: A human inspired approach in air hockey game

In this paper, we present an approach, inspired by human behavior, in predicting the state of a high speed object based on the state of another object that causes such a high speed, e.g. predicting the state of a high speed puck which is hit by a lower speed paddle in the air hockey game. The proposed approach eliminates the need for high speed sensors, such as high speed cameras and grabbers, for capturing and processing high speed motions. This approach has been implemented and tested on an air hockey simulator and in real images taken by an off-the-shelf camera. The results show that the low speed of a paddle in the air hockey can be easily captured by an off-the-shelf camera and then the state of the puck, which is the speed and its pose, after being hit by the paddle can be calculated.

[1]  Bernhard Nebel,et al.  KiRo - An Autonomous Table Soccer Player , 2002, RoboCup.

[2]  Christopher G. Atkeson,et al.  Learning from observation using primitives , 2001, Proceedings 2001 ICRA. IEEE International Conference on Robotics and Automation (Cat. No.01CH37164).

[3]  Mark W. Spong,et al.  Control of Planar Rigid Body Sliding with Impacts and Friction , 2000, Int. J. Robotics Res..

[4]  Mark W. Spong,et al.  Neural network-based state prediction for strategy planning of an air hockey robot , 2001, J. Field Robotics.

[5]  A. Ishlinsky.  On the dynamics of a system of rigid bodies , 1973 .

[6]  Hiroshi Noborio,et al.  Computing and calibrating collision impulses and its application for air hockey game , 2002, IEEE/RSJ International Conference on Intelligent Robots and Systems.

[7]  Wen-June Wang,et al.  A vision based air hockey system with fuzzy control , 2002, Proceedings of the International Conference on Control Applications.