AI-Driven Approach for Digital Agriculture: A Case Study on Coffee Leaf Disease

This study presents an AI-driven approach to coffee leaf disease classification by integrating deep learning and computer vision. Using DenseNet, MobileNet, and ResNet with transfer learning and data augmentation, the system achieved high accuracy, with DenseNet consistently performing the best. The experimental results highlight the influence of learning rate and class distribution on model performance. Model conversion ensures efficient deployment and supports real-time diagnosis. The proposed solution bridges AI research and practical agriculture, enhancing disease management while paving the way for scalable adaptive precision farming.