Localisation and monitoring grazing cows using computer vision

This technical article discusses challenges in pasture-based livestock production and introduces a novel outdoor monitoring system using computer vision. This system employs the YOLOv8 model and an Nvidia Jetson Orin Nano to detect and convert cow positions to GPS coordinates. The approach corrects image distortions and optimizes the field of view for accurate GPS estimation. Future developments include extending the algorithm to recognize different behaviors and integrating with remote sensing for improved feed management and animal welfare.

This work has been developed in the context of the XGain project (101060294). XGain fosters a sustainable, balanced, and inclusive development of rural, coastal and urban areas by facilitating access to relevant stakeholders to a comprehensive inventory of smart XG, last-mile connectivity and edge computing solutions, and of related assessment methods. Learn more at xgain-project.eu.