Prediction of the start of parturition in mares using camera image analysis.

This study aimed to detect characteristic changes in the behavior and body surface temperature (BST) of pregnant mares before parturition using camera image analysis. A total of 53 pregnant Thoroughbred mares, which were individually stabled in box stalls, were included in this study. The mares were continuously monitored using a digital camera system comprising a thermal imaging camera and a visible camera. For the automatic determination of the x and y coordinates of each mare, the centroid of the mares was calculated using image processing software. Subsequently, trajectory analysis was performed using image analysis software. Distances moved were calculated using coordinate data. BST was measured using infrared thermography to measure the average BST. The difference between BST and ambient temperature was adopted to account for the influence of ambient temperature. On the day of beginning the expulsion phase (BEP), the distance moved by the pregnant mares increased significantly more than that 2 days before BEP and was the greatest at 2-1 h before BEP. Furthermore, BST at 6 h before BEP tended to increase as the BEP approached. A significant increase in BST was observed at 1-0 h (P < 0.05). 96.2% of the mares showed an increase in either activity or BST, which indicates the benefit of employing multiple parameters.