基于三维应力对跛行奶牛蹄部参数的提取

Parameter Extraction of Lameness Cows Based on 3-Dimensional Reaction Force

奶牛跛行广泛存在于养殖场中。它不仅影响奶牛的正常生理活动,使得奶牛淘汰率升高,而且人工检测会耗费大量的时间。研究中,提出利用地面三维应力测试系统对行走奶牛的支撑相自动识别,同时进行处理判断,从而获取前后蹄部各自完整的支撑相。所用的方法是基于时间和空间,并且结合四肢动物行走动力学支撑相的逻辑顺序收集压力中心(COP)轨迹点。这些COP的簇被识别为完整的或不完整的,且分别被分离为前蹄或后蹄。通过计算机自动识别程序获得前后蹄部完整的支撑相,并且提取主要的6个特征值。样本试验数据显示,该设计可用于奶牛的步态分析、跛行判断。这将实现牧场自动收集数据进行无人监督步态分析。

The cow limping widely exists in the farms,which not only affects the normal physiological activities of dairy cows,makes the elimination rate increased,leads to great economic losses,but also artificial detection will spend a lot of time. Using three-dimensional stress measurement system to identify the supporting phase of the walking cow,and then obtain the complete support phase of the front and back. The method is based on clustering of Centre of Pressure（ COP） trace points over space and time,combined with logical sequencing of stance phases based on the dynamics of quadrupeds walking. These COP clusters were identified as complete or incomplete,then separated into the fore stance phase or hind stance phase. Through the computer automatic recognition program obtained fore and hind complete stance phase,extract the six main characteristic parameters. Experimental data show that the design can be used for cow gait analysis,limp judgment,which will realize pasture to collect data automatically for unsupervised gait analysis.