Maintaining a horse or a horse stable in general is very costly. Monitoring the behaviour of horses continuously is essential to ensure top physical condition of these animals and is a time consuming process especially when it is part of research projects and requires research scholars to keep track of key activities, poses and relate them to health conditions of the horses. Visual monitoring can be extremely expensive in such scenarios and also slows down the pace of research significantly as the researchers have to spend a long duration of time analysing activities frame by frame and recording the data in structured formats.
Deep Learning based Computer Vision engines diligently trained to detect animal poses like Standing, Walking, Sleeping, Foraging, Getting up, Defecating etc. can be employed instead to capture different poses of these animals from video footages and to provide these data in a tabular form. This time series data can then be easily explored by researchers to arrive at informed conclusions about the animal behaviour.