Remote Intelligent Integrated Monitoring System Of Agricultural Diseases And Insect Pests

With the application and popularization of Internet of Things technology,smart agriculture is the only way for the future agricultural development.Compared with developed countries,there are serious deficiencies in the level of agricultural information development in China.The information monitoring of crop growth environment,the monitoring of pests and diseases and the prevention mainly depend on artificial field investigation,but the real-time and automation degree is not high.Some of the systems for information collection are not effective enough for data summary and analysis.It is difficult to achieve data-driven,and many problems hinder the development of agricultural productivity.In view of the present situation of China’s agricultural information service,this thesis improves and realizes a set of remote automation and intelligence based on the original system of laboratory,relying on sensor technology,Internet of Things technology,data encryption technology,server distributed technology,mobile APP software design,Web development technology,database technology,etc.The visualized remote and intelligent integrated monitoring system of agricultural pests and diseases provides a set of highly available solutions for the current construction of agricultural information technology,making up for some of the deficiencies.The system studied in this paper includes three subsystems: information collection,system service and user client.The main research contents are as follows:1.Improve and implement the software and hardware functions of the information collection side.The hardware selection and the main software design of the information collection end have been completed in the previous work.The work of this paper is as follows: First,to improve the software process,add some logical judgments,optimize the logical process,and enhance the system stability.Second,the time-stamp based one-time Password（TOTP）algorithm is combined with AES encryption and decryption algorithm to make the keys between the collector and the server update in real time without transferring.Thirdly,the problem of electromagnetic interference in the process of hardware integration is analyzed and solved.By testing,the appropriate shielding wire is selected to solve the problem of electromagnetic interference,which improves the electromagnetic compatibility of the system and ensures the stable operation of the system.2.Redesign and implementation of the system server.Firstly,a highly available and distributed message queue structure is introduced.The improved system server consists of three parts: the collector server,the client server and the message queue between them.They serve the client,the collector and the two servers respectively.First,the server on the collection side is designed and implemented,and a long connection with the collection side is established and the message mapping from the collection side to the message queue is implemented.Secondly,the original client server is improved to implement client-to-message queue instruction mapping,to provide different time-granularity query functions for the client interface functions,parameter resolution functions,database query functions,and so on.3.Design and improve the system client software.First,combined with the interface functions implemented by the client server,different time-granularity data query and chart display functions are added to the Web client.Secondly,the i OS App based on Objective-C is designed and implemented,which implements user login,device bind/unbind,parameter settings,video viewing and cloud control,data query and chart display.4.A leaf disease image detection algorithm based on Res2 Net convolution module and YOLOv3 network was designed and implemented.The training and testing of Grape Black Rot leaf image were carried out,and the trained model was deployed on the server of this system.It ensures high efficiency with high detection accuracy and meets the system requirements.