Design And Implementation Of Online Recognition System For Growing Tendency Of Vegetable Heart Based On Deep Learning

Vegetable plays an important role in the field of agricultural planting in China..The quality of its planting process directly affects the income of farmers,and the efficiency of growth discrimination is related to the yield and quality of fruits and vegetables.Although the traditional manual discrimination can get results,it is inefficient and has some subjective factors.Therefore,a simple and efficient method is needed to replace the traditional manual discrimination.In this paper,a set of on-line identification system for growing trend of Chinese cabbage is developed based on the growing trend recognition of Chinese cabbage.The main contents are as follows:(1)Establish data set and data preprocessing.The data set was collected in the planting base of cabbage.By referring to the data and combining with the local agricultural documents,the judging standard of the growth of the cabbage heart was established According to this standard,the cabbage can be divided into four categories:normal cabbage at seedling stage,abnormal cabbage at seedling stage,normal cabbage at growing stage and abnormal cabbage at growing stage,with a total of 1600 image data.Data preprocessing is carried out from two aspects of normalization and data enhancement to provide data support for this paper.(2)The convolution neural network is constructed to identify the growing potential of cabbage heart.Based on the structure of alexnet network,this paper improves it by using small convolution kernel stack instead of large convolution kernel,and adds two new techniques,dynamic learning rate decay and momentum gradient decline,to further optimize the model.Set up pre experiment to test the rationality and superiority of the model.Finally,the accuracy of 93.19%is obtained,and compared with the alexnet model before the improvement,the convergence speed is significantly improved.(3)Based on Java Web technology,an online recognition system is built to realize the whole process from uploading pictures to displaying results.The system uses Tomcat as the server,using HTML,CSS,JavaScript three main technologies to complete the front-end page of the system;using Java's built-in method to call Python files inside the server,so as to realize the model running inside the server;through asynchronous communication technology,the identification results are transmitted back to the front-end page.In this paper,deep learning technology is introduced into the web platform,and a complete online recognition system of cabbage growing trend is developed.The system has a high recognition accuracy and good stability,and realizes the purpose of identifying cabbage growing trend accurately and timely.