A METHOD OF SIGNAL-PROCESSING IN MOTION ANALYSIS OF THE TROTTING HORSE

The aim of this paper is to present a method of signal processing nece ssary for motion analysis in the trotting horse. Motion analysis is wi dely used to assess lameness in horses. By definition, lameness in tro t is present if the movements during the stance phases of both fore or hind limbs differ. The motion of the horse is recorded using a system for motion analysis (Selcom, 1983, SELSPOT II User Manual, Pad Nr. 67 10) and the vertical motion of the head during both stance phases is c ompared. The symmetry is analyzed comparing the values of the Fourier coefficients. Additional head movements are 'single events' and presen t a problem. In this article we propose a mode of data processing usin g a system matched filter that neglects the nonperiodic additional mov ements and allows calculation of the representative Fourier coefficien ts. In comparison with a filtering method previously used in motion an alysis this method is proved to be advantageous, as the precision of s ymmetry comparison increases. Copyright (C) 1996 Elsevier Science Ltd.