Leveraging Artificial Intelligence for Age Prediction in Cattle based on Dentition

Age estimation is of vital importance in cattle for various reasons such as health monitoring, breeding, and economic planning. Conventional methods rely on dentition but they are subjective and require considerable expertise, which can lead to inaccuracies. This research aims to address this problem through the use of artificial intelligence by comparing two approaches. In our first approach, bounding box annotation was performed on cattle teeth, and using a YOLOv8 object detection model, we predicted the age class of cattle. For the second approach, individual teeth were segmented to extract morphological features to train machine learning algorithms. Through these methods, this research provides an innovative solution for age estimation through computer vision, which had not been explored before. A demo video of cattle age estimation using dentition can be found here.