2025 Volume 87 Issue 3 Pages 197-203
Compared to other livestock, the establishment of husbandry techniques in goats has lagged behind and there is a need to develop behaviors monitoring system for goats. In this study, behavioral features were extracted from time-series data measured from an inertial sensor mounted on the goat's neck. We made training dataset from the feature values for each of ruminating, foraging, resting and other which are important behavioral indicators, and developed a classifier that classifies the four behaviors. As a result, characteristic periodic waveforms indicating ruminating were identified in pitch and yaw angular velocities. In addition, the number of ruminations in diseased goats increased about sevenfold after medication. Furthermore, the accuracy of the developed classifier was 81.9 %.
