Proposal and evaluation of behavior estimation method for Shiba goats

Abstract

For the successful mating of female and male mammals, their estrus behavior has significant importance. Professional evaluators observe the goats' behavior from the viewpoint of the management of estrus behavior in female goats. However, this observation method requires extensive time and expense in case the number of goats is more or if an evaluator needs more time for observing goats. Therefore, we tried tracking the movement of female goats and applied a hidden Markov model to the goat's data for estimating whether female goat was estrus. Resultantly, we detected the estrus behavior with high accuracy, thereby validating that using a hidden Markov model for estrus estimation is an effective method.