AI-Driven Farming: CNN-Based Rice Leaf Disease Identification and Classification

This research investigates the performance of three distinct neural network models, including the Convolutional Neural Network (CNN) architecture, InceptionV3, and Xception, for classifying rice plant diseases. Evaluation metrics encompassing accuracy, precision, recall, and F1 score were employed to assess the models' efficacy in classifying diseases (Blast, Blight, and Tungro). The findings reveal that the Xception Transfer Learning Model outperformed the other models, achieving an average accuracy of 0.97. InceptionV3 also exhibited strong performance, with an average accuracy of 0.95. Meanwhile, the CNN Architecture model presented lower performance, with an average accuracy of 0.94. Confusion matrices were analyzed for each model, offering comprehensive insights into classification performance. These results facilitate the selection of the most suitable model for disease classification in rice plants, with the Xception Transfer Learning Model emerging as the preferred choice.