Chinese researchers have developed a lightweight AI model designed to recognise cattle behaviour in video footage captured by quadruped robots on pastures, improving herd management and feeding efficiency. This is reported by
China Daily, a partner of TV BRICS.
The AI model was developed by the Agricultural Information Institute of the Chinese Academy of Agricultural Sciences. The source stated accurate and rapid detection of typical cattle behaviours is critical for disease diagnosis, oestrus monitoring, calving prediction, and overall health assessment.
The invention can identify multiple behaviour patterns in complex environments, making it suitable for real-time deployment on mobile robots.
By integrating multi-scale attention networks, adaptive decomposition and alignment heads, and other advanced techniques, the system addresses challenges such as lighting fluctuations, motion blur, and object occlusion within herds.
The model accurately recognises six typical cattle behaviours, including feeding, resting, walking, and licking, striking an optimal balance between detection accuracy and computational efficiency. The AI model provides essential technical support for the large-scale development of pasture robots.
Photo: ahavelaar /
iStock
Самые
актуальные новости стран БРИКС https://tvbrics.com