An End-to-End Deep Learning Method for Potato Blight Disease Classification Using CNN

This paper proposes an end-to-end approach using deep learning to classify potato blight diseases using Convolutional Neural Networks (CNNs). The system is designed as a user-friendly web-based application and mobile app for farmers to easily predict early and late blight diseases in potato plants. Early detection and accurate identification of these diseases can save significant resources in agricultural practices. The proposed method utilizes deep learning and CNNs to process uploaded images of potato leaves, differentiating between early blight, late blight, and healthy plants with high accuracy. The project involves data collection, model building, MLOps using TensorFlow Serving, backend server development with FastAPI, and deployment on Google Cloud Platform (GCP) with GCP Functions. A React Native mobile app serves as the user interface. Future work includes publishing the application on Google and Apple Play Stores and expanding the model to cover additional plant diseases for a more comprehensive agricultural solution.