Research on crop disease identification based on GoogleNet

Inhalt:
Crop diseases and pests seriously affect the growth and production of crops. Timely and accurate identification and control can effectively improve the yield and quality of crops. A total of 15 types of data in the data set of Plant Village, including healthy pepper and plaque leaves, healthy potato early blight and late blight, and common tomato diseases, were selected for classification and recognition by using deep learning algorithm. The correct rate of crop disease identification based on GoogleNet is 96.86% in the case of small samples. This study has obvious advantages in small sample size and processing efficiency, it can meet the agricultural needs of rice pest classification and detection in the field, and has great application potential in the field of agricultural informatization in the future. It is suitable for field interactive collection and processing combined with mobile devices.