Study on Monitoring Methods of Z of ginseng pests and Diseases Based on Transfer Learning Method

Ginseng pests and diseases have now become one of the important factors affecting the yield and quality of ginseng cultivation. Traditional pest and disease monitoring methods mainly rely on the experience of plant protection personnel and manual identification, which have certain subjectivity and limitations, and are difficult to meet the demand for large-scale and high-efficiency real-time monitoring. The introduction of transfer learning provides a new idea to solve the problem of ginseng pest and disease monitoring in small samples. In this study, we constructed a ginseng pest and disease image dataset and used data enhancement techniques to extend the sample size to alleviate the small sample problem. Using the migration learning method in deep learning, the migration learning method is applied to the pest and disease classification task by fine-tuning the hyperparameters and layers of the ResNet-50 model. The experimental results show that the fine-tuned ResNet-50 model achieves a classification accuracy of 95.01% on the test set, which is significantly better than the traditional method and other pre-trained models, demonstrating the efficient performance of the model in the pest classification task.