Measurement And Test Of Dairy Cattle Body Size Based On Kinect

As an important source of high quality protein,milk has been paid more and more attention in modern life.To ensure the health of dairy cows is the basis to ensure the quality of milk.As one of the important standards to measure the health of dairy cows,the body size of dairy cows has been paid more and more attention in the modern dairy cattle breeding industry.The traditional acquisition of dairy cattle body size requires a lot of manpower and material resources,which can not meet the requirements of large farms to obtain dairy cattle body size quickly,efficiently and accurately.Binocular camera or industrial CCD camera is often used in the widely used dairy cow image acquisition.The limitation of traditional camera operation limits the speed and quality of dairy cow body size acquisition.In view of the disadvantages of traditional dairy cattle body gauge acquisition,modern animal husbandry is looking for a way out in the direction of high technology.The low-cost and powerful Kinect camera has become a research hotspot.Under the support of the special fund "Integrated Monitoring platform of Precision breeding Dairy cattle phenotypic characteristics(Project No: 310171)",this paper studies the acquisition of dairy cattle body size based on Kinect.The main contents are as follows:(1)the scheme of obtaining milk cow's body size based on Kinect was put forward.By referring to domestic and foreign literature and combining with the demand of modern dairy cattle breeding,this paper presents an automatic and stress-free acquisition scheme for dairy cattle body size based on Kinect depth camera.(2)the construction of the platform for obtaining the milk cow's body gauge.The dairy cow depth image acquisition board was built with Kinect as the main body.The scaffold was constructed with truss material,the dairy cow depth image acquisition board was installed,the cow positioning system based on infrared tube sensor and the cow identification system based on ear marker reader were installed.The platform for obtaining milk cow body size is completed.(3)acquisition and preprocessing of dairy cow depth image.The depth images of dairy cattle were obtained and stored in classification.The background subtraction,threshold segmentation,median filtering,open operation,connected domain detection,edge detection and other algorithms are used to obtain the dairy cow target area.(4)obtaining the milk cow's body size.Through convex hull algorithm to remove cow head and tail to obtain cow body area;to screen point cloud data of cow and draw 3D model of cow back;to analyze and calculate the internal corner point of dairy cow target area to obtain the cow body height,body width,body length,body length,body height,body width,body length,body length,body height,body width and body length.Back projection area and back projection volume.(5)Test and data analysis.Establishment of GUI human-computer interface to obtain dairy cattle body size data.The milk cow body size data obtained by manual measurement was used as standard body size data.The two groups of data are compared and analyzed to verify the accuracy of the platform.