Our entry "Knowing Cattle with Wisdom-Automatic Scoring System for Cow Body Condition Based on Deep Vision" uses advanced deep learning technology and integrates three functions of cow target detection, identification and body condition scoring to realize automatic scoring of cow body condition. The system uses YOLOv8-OBB rotating target detection technology to accurately locate the back area of the cow, uses the Multiple Granularity Network to realize the individual identification of the cow, and realizes the high-precision body condition score through the multi-feature weighted calculation model. At the same time, in order to alleviate the pressure of cloud server computing, the edge equipment processes the data and transmits the obtained cow image, identity and scoring data to the cloud platform, providing data storage, analysis and visualization services, providing real-time monitoring and decision support for ranchers to help optimize pasture management. The detection accuracy of the back area of the system is as high as 98.8, the identification accuracy is 97.4, and the body condition scoring accuracy is 98.1 within 0.25 points error. At the same time, the system has the ability to work in real time all day long, providing personalized and accurate health management support for the pasture, significantly improving the management efficiency of the pasture, and helping the intelligent and modern process of China's dairy industry.