Monitoring the daily change in cow's body weight in relation to cow health and behavior

The monitoring of cow’s weight, health and behavior plays an important role in the dairy farm. The goal of this research is to monitor the daily change in cow’s weight without supervision and its relation to health and behavior. The combination of computer vision (to identify the cow and its position using the torso segmentation) and data analysis (to interpret the weight changes) can help create an autonomous monitoring system. This research proposes a method for using video data to determine the reliable cow weight measurements, to synchronize the video data and weight log data, as well as how to interpret weight variations in relation to grazing behavior, milk production, and age. The findings reveal that weight is positively related to the cow's age, that weight changes are likewise positively related to milk production, and that grazing pattern has an impact on weight, with more grazing decreases weight.