跨域感知 | 智慧決策
Cross-Domain Perception | Intelligent Decision-Making
Contact us to learn more.
Intelligent Cattle Lameness Monitoring System
The system employs a deep learning-based cattle lameness classification model built on the lightweight MobileNetV3 architecture, enabling multimodal fusion classification. It automatically identifies lameness severity by analyzing cattle Gait Energy Images (GEI) and History Energy Images (HEI).
Application of weight prediction for Holstein dairy cows in non-pregnant and postpartum stages
A non-contact weight prediction system for Holstein dairy cows was developed based on depth sensing technology, designed to predict weight changes during non-pregnant and postpartum stages. The system utilises an Intel RealSense D455 depth camera to capture depth image information from cow’s dorsal, hips, and side regions, extracting effective body surface feature data through a systematic data processing workflow.
Analysis of Heat Stress in Dairy Cattle
Thermal image sequence analysis automatically extracts nasal temperature and respiration rate to build a heat stress severity classification model. This non-contact monitoring eliminates measurement-induced stress on cattle, providing authentic physiological assessments. Farm managers can detect early heat stress signs and implement cooling measures promptly, reducing negative impacts on cattle health, milk production, and reproductive efficiency.