Research On Remote Monitoring System Of Thrips Tabaci Based On Image Processing

Thrips tabaci is the main pest of agriculture in China.Tobacco thrips mainly damage crops such as vegetables and flowers.Most plants harmed by thrips are closely related to human life and health.Furthermore,thrips damage can also cause crop growth.Slowly,it affects the quality of crops,resulting in a large reduction in crop yields.Therefore,thrips pests need to be monitored.At present,the monitoring of thrips pests in China mainly relies on the trap board to trap and capture thrips pests,and then the plant protection personnel exclude non-primary pests,manually identify and count thrips pests,and finally,the thrips count results are used to monitor the field The population dynamics of thrips pests.At present,the monitoring of thrips pests in China is still at the stage of artificial identification of pests,which has the problems of low efficiency,high labor intensity and non-timeliness.To this end,this paper studies the image monitoring system of tobacco thrips pests based on image processing.The system consists of an image acquisition platform,a data storage server and a processing terminal.Using this system,it is possible to realize image collection of thrips pests,environmental data collection,thrips pest counting,and view thrips images and e nvironmental data.On the one hand,this research is the basis of a remote monitoring system for field pests and promotes the development of plant protection information technology.On the other hand,it provides a reference and basis for identifying tiny insects using image processing.The main findings are as follows:(1)The trapping equipment for thrips pests consists of triangular traps,blue sticky insect plates,thrips trapping core,and insect nets.The trapping equipment for thrips pests collects thrips pests in the greenhouse.The thrips core ensures that only thrips pests in the greenhouse are trapped,and the insect net is used to prevent other unrelated larger pests from entering the trap to ensure the unity of thrips pests.(2)The image acquisition equipment consists of a raspberry pie,a wide-angle fisheye camera,a temperature and humidity sensor,and a wireless network card.The image acquisition device is responsible for driving the high-definition camera to take pictures of thrips that are trapped.The computer vision library OpenCV and image processing technology are used to calculate the number of thrips on the blue stick armyworm.Among them,the image processing technology first looks for the best color channel and K-mean mean clustering method is used to segment the blue sticky insect plate image from the black background image,and the segmented blue sticky insect plate image is subjected to Gaussian filter to denoise and grayscale,and the grayscale histogram is used to convert the grayscale image Perform binarization processing,use image opening operation to remove the noise of the binarized image,which is convenient for separating the stuck objects,use image edge detection to obtain the image contour,and obtain the image contour,obtain the image within the contour,use training The VGG19 network completes the identification and counting of thrips images.Among them,the results of the image recognition count and the manual counting of thrips are about the same.The maximum value of the relative error between the two is 8%.Therefore,the image counting method can be applied to thistle Identification and counting of horse pests.(3)The data transmission platform is composed of the ECS cloud server launched by Alibaba Cloud and the OSS object storage server launched by Qiniu Cloud,and adopts a separate communication and data transmission model.The ECS cloud server is responsible as an intermediary between the image acquisition platform and the client to complete the information transmission between the image acquisition device and the client.The OSS cloud database is responsible for storing the temperature and humidity data of thrip pests collected by taking pictures,and backing up the pictures and data to ensure the safety and e ase of use of the data.(4)The image acquisition platform,data transmission platform and client use TCP / IP protocol for network connection,and the open source high-performance reverse proxy application frp is used to complete intranet penetration,making the image acquisition device connected to the local area network Raspberry Pi can provide services to the external network,so that the client can remotely control the Raspberry Pi through the network.By sending parameter commands on the client to realize the regular photo capture function of the image acquisition device,it can also enable the Raspberry Pi to build a thrip Web page Public network connection and viewing.In short,this experiment uses a combination of computer vision image processing technology and an embedded system to study the image acquisition and processing system of thrips pests.Thrips pest monitoring.