Design Of Monitoring System And Identification Technology Of Rumination Behavior Of Sheep

With the development of society and the progress of science and technology,the traditional grazing breeding method has lost its original advantages,and large-scale sheep farms have gradually become an important part of the mutton industry.With the great advantages brought by the development of standardized breeding,large-scale farms not only improve the income,but also reduce the environmental pollution.However,captive sheep have relatively weak ability to resist diseases.If the sick sheep can not be found in time,it will bring great potential risks to large-scale breeding,which will seriously restrict the development of mutton industry.The best way to solve this problem is how to be found in time in the early stage of sheep disease and avoid serious or spread of the disease.As one of the representatives of ruminants,sheep’s rumination behavior contains physiological,health and other information.If the rumination behavior can be accurately obtained,it will have important practical significance for the analysis of sheep’s health level.In order to monitor the rumination behavior of sheep and reduce the impact on the daily behavior of sheep,this study carried out the research on the rumination behavior monitoring system of sheep based on computer vision technology.The main research contents include the design of monitoring system and the acquisition of sheep rumination data.First,design and optimize the monitoring layout to solve the problem that the farm cannot accurately obtain the daily behavior data of sheep,and put forward the full coverage monitoring layout under the optimization of monitoring effect,so that the monitoring rate of sheep can reach 78%,which lays a foundation for obtaining the rumination data of sheep.Secondly,in order to accurately obtain sheep rumination data,this thesis proposes a method to extract sheep mouth region based on Gunnar farneback dense optical flow algorithm.Using this method,the optical flow vector generated by the movement of all pixels in each frame in the video stream in the time domain is calculated,and the vector is segmented to obtain the image data of the change of the mouth area when the sheep ruminate;Subsequently,through the mouth width detection algorithm,the change law of mouth width during rumination was obtained,and then the chewing information and rumination data of sheep were obtained;At the same time,by marking the head position of sheep,the rumination data of multi-target sheep are obtained;Finally,the feasibility and accuracy of the system were verified by comparing the experimental and artificial observation results,which provided a basis for the further study of sheep rumination behavior.It is a simple and accurate method to obtain ruminant data by studying the change of ruminant mouth width,and non-contact monitoring is more suitable for the requirements of animal welfare breeding,which points out the direction for the future ruminant behavior monitoring research.