Design And Implementation Of Intelligent Detection System For Cherry Pests And Diseases

The cherry,also known as "Peach-containing",is the earliest mature tree species among deciduous fruit trees.Its fruit is not only crystal but also full,and the pulp is rich in nutrients,which is very popular among consumers.However,the planting process can lead to a variety of pests and diseases,which can easily cause a decline in production and low economic efficiency.Traditional detection of pests and diseases mostly relies on growers’ experience and agricultural institutions’ professional testing,which is time-consuming,labour-intensive and has poor real-time performance,making it difficult to meet the development needs of modern agriculture.Therefore,it is essential to achieve accurate and rapid detection of cherry pests and diseases to minimize yield and economic losses.Based on the above background,relying on the cherry intelligent planting project,this paper designs and develops an intelligent detection system for cherry pests and diseases,which realizes intelligent image recognition as well as rapid diagnosis of cherry pests and diseases.The main research contents,methods and results of this paper are summarized as follows:(1)Establish a knowledge base of cherry pests and diseases.Through crawling techniques,consulting books and on-site research at the base,we collected detailed introduction,types,symptoms,main damage parts,disease patterns,control measures and pictures of cherry pests and diseases.And the corresponding database was constructed using My SQL and Neo4 j database tools.(2)Research the image recognition algorithm of cherry pests and diseases.In this paper,aiming at 28 kinds of cherry pests and diseases,based on data enhancement technology,three popular convolutional neural network models,VGG16,Inception V3 and Res Net101,are finetuned by transfer learning method.Finally the Res Net101 model with the best recognition results for cherry pests and diseases was selected by comparing the accuracy of the models to provide algorithmic support for the subsequent system development.(3)An intelligent detection system for cherry pests and diseases was developed according to the actual needs of the project.The system includes two parts: the We Chat applet terminal and the web terminal.By applying the algorithm model trained in(2),it finally realizes the functions of cherry pests and diseases image recognition,diagnosis based on knowledge graph,multifaceted data display of cherries and comprehensive management of various cherry cultivation data.Through system testing and algorithm experiment verification,the intelligent detection system for cherry pests and diseases implemented in this paper meets the design requirements as well as provides convenient and fast detection for cherry growers,which can effectively improve cherry yields and reduce economic losses.Although the main object of this paper is the cherry,the model and method can be extended to most other crops in the intelligent planting process,which has certain application value in improving agricultural efficiency.