Research on Intelligent Recognition Solution of Tobacco Disease on Android Platform

Abstract

In order to improve the recognition accuracy of tobacco diseases, improve the recognition efficiency and convenience, and reduce the recognition cost, this project carried out the research on the recognition technology of tobacco diseases based on deep learning. First, the data set was established. The data set is consisted of several kinds of common tobacco diseases images which were labeled according to the experts’ diagnosis results. Second, the YOLOv7 network model was studied and pruned considering the recognition rate and accurate. Third, the pruned model was trained using the established training dataset. Then, the trained model is ported to Android system. Finally, an experimental testing was carried out, and the results show that the model can run efficiently in Android system with the detection accuracy above 90%.