Three-dimensional motion measurement of golf putter head and estimation of putting results by neural network

Abstract

This study was conducted to clarify relationship between putter head motions of golf and ball hitting direction, distance with use of statistical techniques. Measurement was carried out by measuring systems for golf putting which composed of 3D gyro sensor and infrared sensors. From the measurement data, 27 parameters related to putting were selected. From the 27 parameters, 10 parameters related to hitting distance and direction were selected by analysis of variance. 10 parameters were applied principal component analysis to eliminate data redundancy. Using the first to forth principal components obtained by principal component analysis, the hitting distance and direction were learned and identified by the neural network. In addition, it was shown from the principal component analysis, angle around y axis, z axis at impact, angle about z axis, maximum of angular velocity, angular velocity around z axis at impact,the maximum and minimum angles around the y axis are important parameter for expressing the ball hitting result. Moreover it was shown that the parameter that most influence the estimation of the hit ballistic result are the angle around the y axis at the time of impact, and the angular velocity around the z axis at the time of impact.

Keywords

• sports engineering
• canonical discriminant analysis
• golf
• putter head motions
• neural network