A Comprehensive Study on Paddy Leaf Disease Detection using CNN and Random Forest

In the context of paddy leaf diseases, this study thoroughly assesses disease classification performance. Brown Spot Disease, Blasting Disease, Sheath Blight, Bacterial Leaf Blight, and Leaf Scald are only a few disease types thoroughly evaluated in this study using essential metrics, including Precision, Recall, and F1-Score. Brown Spot Disease achieved a 95.24% Precision value, demonstrating the model's accuracy in making correct predictions. Leaf Scald's Recall value is 93.65%, demonstrating the model's success in identifying true positives. Sheath Blight's F1-Score of 92.52% exemplifies the study's model's all-around success. These results are supported by accuracy rates between 0.96 and 0.98, demonstrating the model's superiority in diagnosing diseases. Class support and proportion numbers are also emphasized for their importance in the study, as they provide crucial context to the investigation by measuring the number of examples in each class and their corresponding representation throughout the dataset. The study demonstrates the model's promise for actual disease management and yield optimization in rice farming, providing a trustworthy instrument for precisely categorizing various paddy leaf diseases. The model's impressive 92.194% accuracy indicates its prowess in categorizing a wide range of data.