Research And Realization Of Apple Leaf Disease Identification Based On Computer Vision Technology

With the continuous development of the information age,the application of computer vision technology to various industries is more and more in-depth,especially in the field of agriculture has been widely used.The rise of the tide of the information age makes the technical model of the combination of the Internet and agriculture has been developed by leaps and bounds.In recent years,with the booming development of apple industry,apple is increasingly affected by diseases in the process of growth,which will harm the yield and quality of apple in serious cases.In the current stage of planting,it is an important method for fruit growers to effectively identify and prevent diseases in advance,so as to avoid economic losses caused by diseases.However,at present,fruit growers are unable to identify diseases timely and accurately only by relying on their own experience.Because of the high similarity of apple leaf disease images,it is difficult for the traditional learning algorithm to achieve high accuracy.Therefore,this paper studies a kind of apple leaf damage identification method based on computer vision technology,image preprocessing,the deep learning algorithms to identify apple leaf disease detection,build the depth of the residual identification model,learning algorithm of apple leaf disease of plant disease image of apple high-dimensional feature extracting,effective classification of plant disease image recognition for apple.The main contents of this paper are:1.Studied and analyzed the status quo and significance of plant disease recognition technology.Based on the framework of computer vision technology,deeply discussed the deep learning algorithm and residual network structure;2.The image characteristics of apple leaf diseases were analyzed,and four major diseases,namely,spotted leaf disease,Mosaic disease,gray spot disease and rust disease,were obtained.At the same time,the disease images were preprocessed.3.Build an apple leaf disease recognition model based on residual network;4.Construct neural network models of different depths for model training,identify and detect the diseases of apple leaves with the obtained models,compare the accuracy of the models,and discuss the optimal disease recognition method;5.Designed and implemented an apple leaf disease recognition APP.The study found that,compared with the traditional neural network model based on the residual network structure of the apple leaf disease recognition model of identifying the highest precision,accuracy is 92%,can be good for detecting disease identification,the study of the apple leaf diseases provides the technical support,use of training good apple leaf disease recognition model,designed and developed a simple apple leaf disease recognition APP,can realize the function of the APP,has a good practicability.