Automatic identification system of pigs with suspected case based on behavior monitoring

An automatic detection method of pigs with suspected case was proposed after analyzing the disadvantages of traditional observation methods. Based on ARM platform the embedded system was designed to monitor the excretion behavior of pigs behavior in 24 hours. When the abnormal behavior detected by the moving object detection and symmetrical pixel block image recognition algorithms took place, the relevant pig would be regarded as the suspected case and the corresponding image would be sent to the surveillance center through GPRS networks. The experiment results for 10 Yorkshire pigs showed that the detection accuracy of suspected case is 78.38%. The method and monitoring system will be helpful for improving production automation in modern pig farm.