基于图像和声音技术的种鸡舍内异常事件监测方法

针对种鸡养殖规模不断扩大,饲养员无法靠人工方式实现鸡舍24h连续监测并及时发现舍内异常状况的问题,提出一种非接触式、24h连续、自动化的监测方法,采用Kinect设备同步采集图像和声音数据,基于LabVIEW软件分析并预警鸡舍内的异常事件。该方法将图像和声音技术相结合,获取种鸡群体日平均活动指数和声音能量,对禽舍内影响种鸡生产性能的异常事件进行预警研究。结果表明,该方法预警异常事件的准确率为73.9%,漏诊率0,误诊率26.1%,相比于单一图像或声音监测手段,该方法能够较全面地监测影响种鸡生产性能的异常事件,为深入研究动物行为和动物福利提供了一种有效的参考手段。

A method based on image and sound processing for monitoring abnormal events in a breeder house

Because of the incremental scale of the farms and the corresponding higher number of breeding hens, it is increasingly difficult for farmers to monitor their animals over continuous 24 hours and to discover their abnormal situation timely. In this study, a type of contactless, 24-h continuous and automatic monitoring method was proposed by using Kinect synchronously to acquire images and sound data. Then, the data were analyzed through the LabVIEW software for the early warning of abnormal events occurred in the henhouse. This method combined image processing and sound processing to calculate daily activity index and sound energy for early warning study of abnormal events affecting hens'' production performance. The results showed that this system reported abnormal events with an accuracy of 73.9%, the rate of missing abnormal events (P_FNR) of 0, the rate of misjudging abnormal events (P_FPR) of 26.1%. Compared with single image analysis or sound analysis method, this method can fully monitor abnormal events in a henhouse. In conclusion, this method can provide an effective tool for deep animal behavior study and animal welfare in the future.