Leaf Disease Detection Using Convolution Neural Network with RESNET50

This project presents an advanced leaf disease detection and classification system leveraging Convolutional Neural Networks (CNNs) with the ResNet50 architecture. The system preprocesses input leaf images by standardizing resolution and applying median filtering to remove noise, ensuring optimal image quality for analysis. A fine-tuned ResNet50 model, pre-trained on extensive datasets, is employed to classify leaves as normal or abnormal. For abnormal cases, segmentation techniques are applied to isolate and identify affected regions, improving the precision of disease detection. The model is further trained to classify specific diseases and assess their progression stages. To enhance usability, the system automates email notifications to registered users, providing detailed reports on the detected disease, its stage, and recommended pesticide treatments. By integrating machine learning and image processing techniques, this project delivers a seamless and efficient workflow for real-time diagnostics, enabling proactive disease management and improved agricultural productivity.