Automated Crop Disease Detection and Classification using Deep Learning Techniques

Plant disease detection is crucial for ensuring food security and agricultural sustainability by enabling timely intervention to prevent crop losses. Recent advancements in machine learning and image processing have transformed plant pathology, offering efficient and accurate disease identification. This work involves collecting and pre-processing diverse plant image datasets, which are then resized and augmented using the FastAI library in Python to improve model generalization. A convolutional neural network (CNN) based on the ResNet architecture is trained via transfer learning to classify plant images as healthy or diseased. The trained model achieves high accuracy, establishing a reliable baseline for disease detection. It is integrated into a user-friendly interface with Tkinter, allowing users to classify plant images in real time via webcam or uploaded images. The system provides instant predictions on plant health and disease probability, with keyboard shortcuts for swift assessment of live camera feeds, enhancing decisionmaking for agricultural practitioners.