Authors

Abstract

With the development of modern breeding industry, it is very important to count sheep accurately. In the past, herdsmen used manual statistics to count and manage sheep, which was time-consuming, laborious and often had large errors. In recent years, machine learning methods are widely used in automatic target recognition, which can replace manual labor. This system is based on YOLOv5 algorithm for sheep counting management. The counting of sheep was controlled by two - way counting. This improves the accuracy of counting, saves a lot of manpower and material resources for herdsmen, and greatly promotes the management of animal husbandry.

Abstract in Chinese

在大面积的养殖牧场，随着现代养殖业的不断扩大，如何对羊群进行精确计数显得尤为重要。在过去，牧民们用人工统计的方法对羊群进行计数管理，该方法既费时又费力，还经常存在较大的误差。近年来，机器学习的方法大量用于目标自动识别，可以代替人工劳动。本系统基于YOLOv5算法对羊群进行计数管理，通过双向撞线计数法，采用两条线控制羊的计数，提高了计数精确度，为牧民节省了大量的人力、物力资源，极大地推动了养殖牧业的管理。