作者：段玉瑶 马丽 刘刚remotemonitoringwirelesssensornetworkscompute
摘要：In intensive and large-scale livestock breeding,how to evaluate welfare of each individual effectively has become one hotspot of welfare breeding.In this study,a remote monitor system based on computer vision technology and wireless sensor networks was proposed.In the system,pig videos in the piggery were captured by IP cameras and transmitted remotely to PC systems for objects tracking and trajectory extraction,and the piggery environmental data such as temperature,humidity and harmful gases were collected by wireless sensor networks(WSN)nodes.Traditional Camshift tracking algorithm was improved.The average root mean square errors of traditional and improved algorithm were 78.07 and 32.29,indicating the improved algorithm reduced error and improved tracking stability.The results of the real-time monitoring of environmental parameters by WSN showed that the hydrogen and ammonia concentrations,and the ambient temperature were all kept in the normal range,while the carbon dioxide concentration was exceeded,and the air humidity was slightly high.Through data mining,precise welfare monitoring and breeding management were realized.
注：因版权方要求，不能公开全文，如需全文，请咨询杂志社
