Smart Agriculture: Identifying Plant Leaf Diseases with Machine Learning

Plant leaf diseases cause significant impacts on agricultural production, hence earlier intervention is more effective with effective detection. In this work, machine learning is employed to detect five major tomato leaf diseases, including Mosaic Virus, Late Blight, Bacterial Spot, Powdery Mildew, Leaf Mold, and Tomato Training and model testing were done based on a dataset of 1,485 images. Classification accuracy demonstrates good performance with Precision, Recall, and F1-Score at 94.98% to 95.47%. The Support Ratio in different classes of diseases varied between 0.19 and 0.21, and the model, in every class of diseases, also possessed a similar accuracy of 98%. Moreover, an elaborate study of the classification performance vis-a-vis true positives, true negatives, false positives and false negatives proves how well the model can potentially reduce the errors in classification. The overall accuracy of the model is found to be 95.28%, suggesting that it is valid and efficient as an automated tool for disease identification. This system offers prompt and precise disease identification, allowing farmers to act proactively against crop damage. Utilizing machine learning methods, this study facilitates sustainable agriculture by better monitoring crop health and managing diseases, which ultimately leads to increased agricultural productivity.