Design And Implementation Of Edge Computing Based Detection Method And Prevention Management System For Pine Wood Nematode Disease

As we all know,forests play an important role in ecological environment protection.China is poor in forest resources,which forest area only accounts for 4% of the world.In China,the main tree species is the pine.Due to various factors in the growth,the pine trees are prone to many insect pests and disease.And the pine wood nematode disease is the most harmful,which has spread quickly and caused billions of economic losses.Now it has a tendency to spread to the key forest areas and it’s urgent to prevent and control the pine wood nematode disease.In the census of epidemic trees,the traditional manual inspection method has a huge workload and the forest terrain is rugged and difficult to detect.Therefore,the most suitable method is to use drones to collect images and transmit them to the data center to identify the images.However,the transmission of massive data and re-identification makes it not timesensitive and causes a waste of resources in the computing center.Edge computing is precisely proposed to solve these problems.In summary,this research uses an airborne edge computing platform and a lightweight deep learning model to develop an integrated pine wood nematode disease management system.The specific content is as follows:(1)The Edge Computing Module for Fast detectionThis study proposed YOLOV4-Tiny-3Layer to lighten the target detection model and apply it to the airborne edge computing platform.After the collected images are detected,the suspected disease information is transmitted to the data center for fine-grained detection.And the final result is stored in the My SQL database through python calling the backend API.Experiments show that this method achieves the effects of low missed detection and high accuracy.(2)The Data Collection APP for Prevention and ControlA supporting APP is developed to timely remove and treat the infected trees after double detection by the edge computing platform and the data center,and avoid the spread of the epidemic.This APP would read the information of the diseased trees stored in the database,display its geographic location on the map to facilitate the construction team to carry out related removal work,and upload the corresponding information of the diseased trees after removal to update the database,which is effective to realize the cleanup and supervision of the diseased trees.(3)The Operation Management Web for Prevention and ControlBy using Java,Spring Boot,My SQL,this study developed the web of prevention and control operation management system for pine wood nematode disease.The web has established several major functional modules: "User Management","Authority Management","Pesticide Information Management",and "UAV Image Management",which helps to solve lots of problems,such as the high cost of data sorting time,the untimely control of the epidemic situation and so on.Also,the web will follow up the implementation of curative measures.