Scalable, Autonomous and Continuous AI-based Video Monitoring to Improve Animal Welfare, Increase Performance and Reduce Carbon Footprint

可扩展、自主和连续的基于人工智能的视频监控可改善动物福利、提高性能并减少碳足迹

• 批准号: 10053986
• 金额: $ 148.27万
• 依托单位: CATTLE EYE LTD
• 依托单位国家: 英国
• 项目类别: EU-Funded
• 财政年份: 2022
• 资助国家: 英国
• 起止时间: 2022 至 无数据
• 项目状态: 未结题
• 来源: https://gtr.ukri.org/projects?ref=10053986
• 关键词: Scalable; Autonomous; Continuous; AI; based

DAIRYSCORE will deliver a scalable, autonomous and continuous AI-based machine vision solution for the early detection of lameness in dairy cows. It will feed live footage of cows from on-farm cameras into a video analytics container, extract knowledge from observations using advanced neural networks and calculate individual mobility scores for every cow in a herd. The results are made available in form of actionable insights to the dairy farmer through a web-based dashboard showing how many cows are currently lame or show symptoms of reduced mobility that might develop into chronic lameness. Time series data on individual cows enable farmers to treat affected animals early, prevent the onset of full lameness and reduce the severity and duration of illness. As lame animals are in pain and have a less efficient metabolism leading to lower milk production, early treatment also improves farm productivity and lowers the greenhouse gas emissions per litre of milk produced.

DAIRYSCORE将提供一个可扩展的、自主的和连续的基于人工智能的机器视觉解决方案，用于早期检测奶牛的跛行。它将把来自农场摄像头的奶牛的实时镜头输入视频分析容器，使用先进的神经网络从观察中提取知识，并计算牛群中每头奶牛的个体流动性分数。结果以可操作的洞察力的形式提供给奶农，通过基于网络的仪表板显示目前有多少奶牛跛行或表现出可能发展为慢性跛行的机动性降低的症状。关于每头奶牛的时间序列数据使农民能够及早治疗受影响的动物，防止出现完全跛行，并降低疾病的严重程度和持续时间。由于跛行动物处于疼痛中，新陈代谢效率较低，导致牛奶产量下降，早期治疗还可以提高农场生产率，降低每升牛奶的温室气体排放量。