Leveraging ResNet-Based Convolutional Neural Networks for Early Detection of Corn Leaf Diseases

Timely and accurately identifying corn leaf diseases is essential for minimizing crop losses and supporting global food security. The proposed study introduces a ResNet-50-based convolutional neural network (CNN) model designed to detect and classify corn leaf diseases, including Blight, Common Rust, and Gray Leaf, at an early stage. Leveraging an extensive dataset with diverse visual features, the ResNet-50 model achieves a training accuracy of 98% and a testing accuracy of 97%, with precision values as high as 99% for categories like Common Rust and Healthy Leaves. Key performance metrics demonstrate the model’s effectiveness, showing high recall and F1 scores, which indicate a well-balanced sensitivity and specificity. Comparative evaluations with existing methods reveal that the proposed model excels in accuracy, achieving an overall classification accuracy of 97%, and enhances computational efficiency, reducing processing time to just 50 seconds. Integrating the proposed model into existing agricultural frameworks could substantially aid farmers and agronomists in proactively identifying and managing diseases, potentially safeguarding yields. These findings underscore the promise of the ResNet-50 CNN model as a robust, practical tool that can contribute to more resilient agricultural practices and effective disease management strategies on a large scale.