The Visual Intelligent Foaling-Alarm
Why Use Computer Vision In A Foaling-Alarm?
The process of foaling can be stressful for both the mare and the owner. We wanted to develop a solution that gave the owner peace of mind, without needing to attach anything to the horse.
With Computer Vision we made a product that can interpret from a live video using Computer Vision and Machine Learning.
*Video: The heat-map shows where the Machine Learning Algorithm focus when finding the main patterns of interest.
How Does the Foaling Alarm Work?
The solution has 2 tasks, one is interpretation when foaling has started and one is estimating when foaling is likely. To estimate when fooling is likely we have had extensive research and made a combination of acts that indicate foaling is imminent.
We build it as a classification-model using Machine Learning and embedded it on a raspberry pi. This resulted in a precise and robust solution that can handle to be used in stables.
Where Is It Implemented?
We devolved VIFA as an inhouse project after being contacted by a concerned horse-owner. After hearing about the stressful process of foaling, we had to make this product.
The product has just been developed and can for now only be bought on vifa.ai, but it is the plan to get multiple distribution channels
*The picture is of the first prototype.