Research On Intelligent Identification And Application Of Vegetable Image Based On Deep Learning

With the development of technology,the application of artificial intelligence is getting wider.At present,in the process of purchasing,weighing and settlement of vegetables in supermarkets of China,manual participation is needed at all times.And the agricultural products have excellent quality and huge output,which also poses a serious challenge to the market management of agricultural products.During the trading process of vegetables,merchants still use traditional artificial sorting methods.The mode of operation consumes a lot of labor,increasing labor intensity,affecting operating costs.In order to improve the situation,a vegetable image recognition system is designed to help farmers and vegetable merchants detect vegetables,and then complete the tasks of classification,weighing and settlement.It can liberate the hands of the working people,and also bring good shopping experience to the customers,which has great practical value.The article build a vegetable image recognition model based on deep learning technology,and develop a vegetable image recognition system for vegetable merchant transactions in the market.The main contents of this paper are as follows:(1)Research the relevant theories of image recognition,analyze the image preprocessing technology,feature extraction theory and the classification algorithm.And then focuse on the research and analysis of two kinds of convolutional neural networks Goog Le Net and Mobile Net V2.(2)Collect and construct the data set of vegetable image,the image is operated simply before image preprocessing.After that,the image processing is carried out for the vegetable image data set,and the data required for model training and testing is expanded.(3)Research the machine learning models of LBP+KNN,LBP+SVM and LBP+DT,the deep learning network models of Alex Net,Mobile Net V2 and VGG.Compare and analyze the models,and the model based on Mobile Net V2 network with better test effect is selected.(4)Apply the model to the development of the system.According to the method of software engineering,the intelligent identification system of vegetable image is developed.The feasibility analysis,requirement analysis,the overall design and detailed design of the system are carried out.After that,the user management function,image recognition function and model management function are realized.Finally,the system is tested,and the test results meet the expectation.