Research On Tomato Disease Recognition System Based On Machine Vision

Garbage classification has emerged as a pressing issue in China and around the world.It is a technical management system for waste disposal that is both efficient and safe.While several provinces in China have published mandatory waste classification policies,the implementation results of waste classification in real life are not optimal due to a lack of understanding of waste classification and information about proper classification.People have been trying to use artificial intelligence to automatically classify garbage in recent years,making it easier to classify garbage intelligently.Intelligent garbage classification is based on image classification and recognition.Image feature extraction is the first step in image classification.Therefore,it is very important to use effective feature extraction algorithm for image classification.Deep learning algorithm has made a great breakthrough in the field of image classification.Finally,this study uses the deep learning approach in the area of machine learning to design a garbage classification system.The following are the main contents of this thesis:(1)Analyze and preprocess the Huawei data set,expand the categories in the data set that have a small number of items,and create a data set suitable for garbage classification research.(2)The deep learning framework pytorch is selected to compare the classification performance of several deep learning classification benchmark models on the garbage image data set through experiments.To improve garbage recognition accuracy,the model with the best classification performance is selected,and the recognition accuracy of the improved model is more than 93%.(3)A garbage classification applet is designed and implemented by deploying the trained model to the cloud,which can identify and classify garbage online.Garbage identification,garbage classification guide,and garbage classification answer are the three modules that make up the system.(4)A simple intelligent garbage classifier is developed using the Raspberry Pi and Arduino to better solve the problem of field classification.The garbage image is collected using the camera,and then the image recognition model deployed on Raspberry Pie is used to determine the type of article and then which type of garbage it belongs to.To achieve the desired effect of real classification,the garbage is placed into the corresponding garbage can by rotating slot driven by the motor.