Agriculture plays a fundamental role in sustaining human civilization; however, crop diseases continue to pose a major threat to global food security. Early and accurate detection of plant diseases is essential for reducing crop losses and enhancing agricultural productivity, yet traditional disease identification methods rely heavily on manual inspection by agricultural experts, making the process time-consuming, expensive, and impractical for large-scale farming. Recent advances in deep learning, particularly Convolutional Neural Networks (CNNs), have shown significant potential in automating visual recognition tasks, including plant disease classification from leaf images. This paper presents a CNN-based approach for the automated detection of diseases across multiple crop varieties, including rice, wheat, tomato, and potato. The proposed model is trained and evaluated using the publicly available PlantVillage dataset, which contains thousands of labeled images of healthy and diseased plant leaves. To improve model generalization and minimize overfitting, data augmentation techniques such as rotation, flipping, and zooming are employed. Furthermore, the proposed CNN architecture is benchmarked against well-established models, including VGG16, ResNet50, and MobileNetV2. Experimental results indicate that the proposed model achieves a classification accuracy of 97.4%, outperforming several baseline architectures while maintaining computational efficiency. These findings highlight the strong potential of CNN-based automated disease detection systems for real-world applications in precision agriculture and smart farming environments.
We couldn't load the inline preview. Open the PDF directly instead.
Cited via DOI — this article is permanently archived on Zenodo.