🌿 CropSense AI — Intelligent Crop Disease Diagnostic & Yield Protection Platform

A full-stack AI-driven precision agriculture web application. Farmers and agronomists upload crop leaf imagery to instantly detect plant diseases across 38 classes, calculate potential yield loss, view multi-step chemical/biological treatment roadmaps, and generate exportable PDF diagnostic reports.

---

✨ Features

• 🔬 97.45% Fine-Tuned Disease Classifier: Fine-tuned EfficientNetB3 model trained on 87,900 images across 38 crop & disease classes.
• ⚡ Dual-Model Edge Pre-Check: MobileNetV2 binary classifier for sub-second healthy-vs-diseased filtering.
• 📉 Yield Impact & Economic Loss Regressor: Scikit-Learn RandomForestRegressor estimating % yield reduction and monetary risk based on plot acreage, disease severity, and season.
• 💊 Actionable Treatment & Recovery Roadmap: Step-by-step pesticide/fungicide spray schedule, dosage recommendations, and long-term prevention protocols.
• 📸 Live Webcam & Drag-and-Drop Uploader: Supports camera snapshots in the field or direct image uploads.
• 📚 Multi-Leaf Batch Scanner: Process up to 20 leaf images simultaneously for field-wide disease surveys.
• 📊 Time-Series Health Dashboard: Historical scan tracking with Recharts trend visualizer and SQLite persistence.
• 📄 One-Click PDF Diagnostic Export: Generates agricultural diagnosis sheets via ReportLab.
• 🌐 Bilingual Interface: Instant one-click toggle between English and Hindi (हिन्दी).

---

🏗️ Architecture & Tech Stack

cropsense-ai/
├── backend/                  # FastAPI REST API
│   ├── models/               # Pretrained ML Models & Knowledge Base
│   ├── database/             # SQLite / Supabase Persistence
│   ├── services/             # Classifier, Yield Engine, Recommender, PDF Generator
│   ├── routes/               # /api/predict, /api/history, /api/report
│   ├── main.py               # Application Entrypoint
│   └── requirements.txt
└── frontend/                 # Vite + React + TypeScript + Tailwind CSS
    ├── src/
    │   ├── components/       # ImageUploader, ResultCard, Batch, History, Navbar
    │   ├── lib/              # API Client
    │   ├── App.tsx
    │   └── index.css
    ├── package.json
    └── tailwind.config.js

Technology Highlights:

• Deep Learning / AI: TensorFlow / Keras (EfficientNetB3, MobileNetV2), Scikit-Learn (Random Forest), SBERT (Sentence-Transformers).
• Backend: FastAPI, Uvicorn, Pillow, ReportLab, SQLite3.
• Frontend: React, TypeScript, Tailwind CSS, Lucide Icons, Recharts, Vite.

---

🌾 Supported Crops & Diseases (38 Classes)

• Apple: Apple Scab, Black Rot, Cedar Apple Rust, Healthy
• Blueberry: Healthy
• Cherry: Powdery Mildew, Healthy
• Corn (Maize): Cercospora / Gray Leaf Spot, Common Rust, Northern Leaf Blight, Healthy
• Grape: Black Rot, Esca (Black Measles), Leaf Blight (Isariopsis), Healthy
• Orange / Citrus: Citrus Greening (Huanglongbing)
• Peach: Bacterial Spot, Healthy
• Bell Pepper: Bacterial Spot, Healthy
• Potato: Early Blight, Late Blight, Healthy
• Raspberry: Healthy
• Soybean: Healthy
• Squash: Powdery Mildew
• Strawberry: Leaf Scorch, Healthy
• Tomato: Bacterial Spot, Early Blight, Late Blight, Leaf Mold, Septoria Leaf Spot, Spider Mites, Target Spot, Yellow Leaf Curl Virus, Mosaic Virus, Healthy

---

🚀 Quickstart & Local Setup

1. Clone the Repository

git clone https://github.com/VT-2004/cropsense-ai.git
cd cropsense-ai

2. Start Backend

cd backend
python -m pip install -r requirements.txt
python -m uvicorn main:app --host 127.0.0.1 --port 8000 --reload

Backend runs on http://127.0.0.1:8000 (API Docs: http://127.0.0.1:8000/docs)

3. Start Frontend

cd ../frontend
npm install
npm run dev

Frontend runs on http://127.0.0.1:5173

---

📜 License

Distributed under the MIT License. See LICENSE for details.