Revolutionizing Rice Disease Diagnosis: A Fusionof Convolutional Neural Networks and Support Vector Machines

This study uses a CNN architecture to provide a deep learning strategy for the detection and categorization of eight common rice illnesses. Three layers of convolution, three maximum pooling layers, including two fully linked layers make up the proposed model. The photos of numerous rice diseases were gathered from various sources and included in the dataset for this study. A 2,830 picture-labeled dataset with an 80/20 split between both the testing and training sets is used to train the model. The model that was trained is then assessed using Fl-score metrics for precision, recall, and recall. The evaluation’s findings are shown in a graph where each disease class’s effectiveness is gauged by the percentage of assistance given to each class. According to the experimental findings, the model that was suggested achieves a precision of 81.23%, so it’s comparable to the most recent models. The accuracy of each class is greater than 77%, demonstrating the model’s ability to distinguish between various rice illnesses. showing that the model is capable of recognizing the majority of the instances for each disease class, the accuracy of the recall of every category is also over 55%. Each class’s F1 score is higher than 67%, indicating a decent overall performance for the model. In conclusion, the suggested model has a good level of accuracy, recall, and Fl-score for accurately classifying various rice illnesses. The detection and treatment of rice diseases may benefit from the findings of this research, which will help rice production continue to grow sustainably. It is possible to do additional research to enhance the performance of the model by expanding the dataset and utilizing transfer learning strategies.