An end-to-end full-stack deep learning application that detects diseases in potato leaves with 95–98% accuracy. Built with TensorFlow, FastAPI, React.js, and Docker.
Potato crops are severely affected by Early Blight and Late Blight diseases, causing massive agricultural losses across India. Early detection can save entire harvests. This app allows farmers to upload a photo of a potato leaf and instantly know if it is healthy or diseased.
- Accuracy: 95–98%
- Classes: Early Blight, Late Blight, Healthy
- Model: Convolutional Neural Network (CNN)
- Framework: TensorFlow
- Deep Learning: TensorFlow / Keras
- Backend API: FastAPI
- Frontend: React.js
- Deployment: Docker & Docker Compose
- Model Serving: TensorFlow Serving
PlantVillage Dataset — 2152 potato leaf images across 3 classes.
docker-compose up --build
- Frontend: http://localhost:5173
- Backend API: http://localhost:8000
Manikandan KM — BTech AI & Data Science GitHub: github.com/ManikantanKM16