FloraNet: An Approach to Prediction of Sea Buckthorn Disease based on a Convolutional Neural Network and a Random Forest Method

In the current study, FloraNet was used to present an improved method for diagnosing leaf diseases in Sea Buckthorn. In the proposed model, Convolutional Neural Networks (CNN) and Random Forests (RF) are integrated. Essentially, CNN architecture is composed of a pair of convolutional layers, followed by a max-pooling layer; it is then followed by another pair of convolutional layers, before finishing off with a final pair of convolutional layers. Performance indicators such as Precision, Recall, F1-Score, or Accuracy are used to determine the performance of the model in terms of its performance. There were five disorders under investigation: Phakopsora, Erysiphe, Septoria, Colletotrichum, Phoma, as well as viral infections. Overall, the results show that the Micro Average accuracy for all classes is 95.11 %. According to the results, the individual accuracy ranges from 0.98 to 0.99 for each illness, while the F1-Scores range from 94.02% to 95.91% for each illness. This study concludes that the weighted average accuracy of 95.11 demonstrates a high level of proficiency of the model in identifying diseases, and indicates that the model is capable of identifying a wide variety of diseases. By using this integrated approach, agricultural scientists and researchers are able to detect Sea Buckthorn leaf diseases.