Design Of Automatic Identification System Of Diseased Pigs Based On Behavior Monitoring

The rapid development of the information technology has a profound impact to the traditional livestock breeding way.Precision breeding is being payed more and more attention.Reviewing of relative research at home and abroad and contraposing the time- consuming and laborious disadvantages of the domestic traditional pig behavior observation mode by manpower methods,a diseased behavior automatic monitoring system based on embedded system and machine vision technology is proposed.It is conducive to the timely detection and diagnosis for the diseased or suspected diseased pigs through the image surveillance and identification.It also enables the farming enterprises from economic loss.The system is based on Samsung S3C2410 ARM9 hardware platform,using the embedded monitoring equipment installed in the excretion zones to monitor the pigs' excretion behavior around the whole day.A moving objecting detection algorithm based on inter-frame difference is used firstly to determine whether there is any excretion activities of pigs.The incorrect situation caused by the light and very small objects is considered in the algorithm,which enhances the detection accuracy.No sooner has the excretion activities took place than an image recognization algorithm based on symmetrical pixel block is used to recognize the number on the pig in order to find out the excretion activitie belongs to which pig.The ferture vector of the pig's number is compared with the template one to count the distance between them in order to calculate the results.the algorithm has a low complexity and each processing step saves the embedded hardware resources.Finally,the total number of excretion times of each pig is counted.When a pig's total number exceeds a certain level,it will be regarded as the suspected case and the relevant image will be sent to the surveillance center through the GPRS network of China Mobile.The surveillance center are able to view the real-time images of pigs with suspected case through the browser.PPP dial-up system is used to achieve GPRS modules in the embedded platform for wireless Internet access.And CGI technology is used to achieve remote data access.The experiment was based on a column of 10 Yorkshire pigs in a large-scale pig farms.Experiment results showed that the design method could effectively achieve the automatic detection of pigs with suspected case and the real-time and monitoring effects of the system had been achieved the desired goal.The behavior-based monitoring of diseased pigs automatic identification method proposed by this paper provides the efficient intelligent livestock breeding mode with scientific theory and technical support and possible options.And this also provide our country's intelligent and automatic breeding aquatics industry with a very good experience and some reference values.