The Design And Implemention Of Agricultural Pests Intelligent Monitoring System Base On Internet Of Things

China is a country prone to natural disasters and agricultural pest is one of the main of them. Agricultural pests is a major natural disaster during the process of crop growth, and it can directly affect the yield and quality of crops. The outbreak of agricultural pest is wide range, and has great harm, which can affects our national economy, especially agricultural production. In order to improve the overall level of China's agricultural pest resistance, precise research of how crop growth environment affects the number of plant pest trends is particularly important.This paper first introduced the background and significance of the intelligent monitoring application in agricultural pest based on Internet of things(IOT), analysed the research status and results at home and abroad, and expounded the typical three layer architecture of IOT, including the information perception layer, transport layer and application and management layer. This parper proposed the intelligent monitoring system for agricultural pests based on IOT. Due to the popularity of mobile communication network, the information perception and transmission are integrated for monitoring terminal site. The server platform was the top of system, and it is the focus of this paper.Its main function was to acquire, store, integrate, analyse, and release information in a Standardized and systematic way.Key technologies of server platform were introduced, such as J2 EE, Strut2, My SQL and so on. This paper also introduced the design and development of server, including the function module design, database design. According to the demand analysis and the design goals of the system, the server system included the following modules: user management module, data query module, upload module, download module, data receiving module, equipment control module and warning module and pest diagnosis module. The database included user table, monitoring site data table, pest picture table, device control table and warning information table.Agricultural experts can download the pest pictures to analyse.This paper propose k-means clustering and ellipse fitting method to process the pictures.Firstly, K-means clustering algorithm was introduced to separate the pests body from blade background. Secondly, the least square method was used to ellipse fitting the body. Lastly, ellipticity ratio and area were adopted to distinguish body from impurities. Tests show that this algorithm can count pests number effectively.Eclipse and Tomcat were used as the development platform, My SQL5.6 as database, a variety of HTML, Java Script as design languages, Ajax technology as asynchronous updating technology. The server can store, integrate and analyse a lot of data and pictures. User can also control devices and terminal state and warn disaster through Web site, which finally realizes the automation and intelligence of agricultural regulation.