A5 Robust Measurement of Spin for Pitched Baseball by Automated Image Processing

Abstract

For the understanding of a pitched baseball trajectory, flight angle, spin rate, and spin axis are very important factors which are desired to measure exactly. In a measurement of the baseball spin using a high speed video camera, it is difficult to evaluate exactly through an automated image processing, because of the extremely poor quality of a ball image captured in the outdoors. In order to correct the erroneous results obtained through the automated image processing, a smoothing spline technique is introduced. From the derivatives of the smoothed ball Eulerian angles, an accurate ball spin vector (i.e. spin rate and spin axis) is attained. For the sake of the verification of the automated image processing result, an application program which supports to perform the manual image processing easily and exactly has been developed. Concerning an analysis of the spin vector for pitched baseball images, automated and manual image processing results coincide considerably with each other.