In-situ classification of cattle behavior using accelerometry data

Reza Arablouei, Lachlan Currie, Brano Kusy, Aaron Ingham, Paul L. Greenwood, Greg Bishop-Hurley

Abstract: Highlights•We develop machine-learning models to classify cattle behavior on resource-constrained sensor nodes.•The data is collected by tri-axial accelerometer sensors on collar tags fitted to cattle.•We analyze the statistical and spectral properties of the gathered accelerometry data.•Based on insights gained from the analysis, we extract a small set of interpretable features.•Using the extracted features, we achieve in-situ behavior classification with good accuracy.