Research On Application Of Intelligent Ranch Based On Computer Vision

Grassland animal husbandry,as a traditional industry in China,has faced many challenges in recent years.With the development of large-scale and intensive aquaculture,under the concept of healthy aquaculture,we have found a monitoring plan suitable for modern and large-scale farms,implemented effective monitoring and management of largescale farms,and found problems in the farms in time And taking measures are of great significance.This article is mainly based on video image processing technology and computer vision technology for intelligent pasture application research,improve the farm's information management level,improve the farm's economic efficiency,and promote healthy farming.The application research of intelligent pasture in this paper mainly includes two parts: automatic tracking of cattle and detection of the number of sheep.In this paper,5600 cattle image data sets and 5000 sheep image data sets are produced.Use the YOLOv3 target detection algorithm to detect the position information of cattle and sheep in the image,and use the Deep SORT target tracking algorithm to track the detected target.In the application of automatic tracking of cattle,the detection and tracking algorithm is used to determine the direction of the camera's pan-tilt movement,combined with the PID algorithm to control the movement speed of the pan-tilt camera,to realize the automatic tracking of the camera to a single cattle,and multiple cattle Track the function of a single cow.In the application of sheep number detection,the sheep head target tracking model is established through the detection and tracking algorithm.The two-line counting method is used to select the area with a good tracking effect.The sheep entering the counting area in the sheep channel are started to count and leave the counting area.At the end of counting,the automatic counting function of sheep is realized.The application researched in this paper was tested on Sunite Zuoqi Cooperative Ranch.In the automatic tracking application of cattle,the stability,detection accuracy,tracking effect,testing of cattle tracking is tested.After testing and analysis,the application meets the expected requirements and can realize an automatic cruise when the cattle are not detected.Only cattle automatic tracking,and the function of designated tracking single cattle.In terms of the application of sheep number detection,the video collected by the sheep channel is tested.The experimental results show that sheep detection,tracking,and counting results are good,and can be used as a solution for promotion and application.The application of computer vision technology in the field of intelligent pasture construction has expanded its development in this field,and at the same time has a certain role in promoting the application of computer vision technology in grassland animal husbandry.