IoT Multi‑Factor Smart Door Lock

An advanced IoT door lock system with multi-factor authentication: PIN (4x4 keypad), fingerprint (R307), and facial recognition (Pi Camera), plus a cloud backend (Flask on VPS + Supabase) and Telegram notifications.

🎥 Demo

Click the image above to watch the full demo on YouTube 👆

✨ Features

• 🔒 Triple authentication: PIN + Fingerprint + Face
• 🖥️ Touch-friendly GUI (Pygame) for 7-inch displays
• ☁️ Cloud sync via Supabase + Flask backend (VPS)
• 📱 Telegram notifications (unlock events, failed attempts, admin commands)
• 🔄 Automatic face model retraining and sync
• 📊 Access logs and admin management

🏗️ Architecture Overview

Three layers:

1. Edge device (Raspberry Pi) — GUI, sensors, local auth fallback
2. Cloud backend (VPS + Supabase) — user management, model training/hosting, logs
3. Notifications (Telegram Bot)

🛠️ Hardware Requirements

• Raspberry Pi 4/5 and official power adapter
• 7-inch touchscreen (800x480)
• R307 Optical Fingerprint Sensor (UART)
• Pi Camera Module (v1.3/2)
• SG90 Servo Motor
• 4x4 Matrix Keypad (GPIO)
• Jumper wires, microSD card (≥32GB)

GPIO Pin Connections (BCM)

• Fingerprint (UART): VCC 5V, GND, TX→GPIO15 (RX), RX→GPIO14 (TX)
• Keypad: Rows GPIO 5,6,13,19; Cols GPIO 12,16,20,21
• Servo: Signal GPIO 18; VCC 5V; GND
• Camera: CSI ribbon to Pi camera port

🚀 Quick Start

1. Clone repository

   git clone https://github.com/sasukeuchiha14/fingerprint-camera-door-lock.git
   cd fingerprint-camera-door-lock

2. Raspberry Pi dependencies

   cd "Rasberry pi"
   pip3 install -r requirements.txt

3. Configure environment

   cp ../.env.example .env
   nano .env
   # Fill in BACKEND_URL, Supabase keys, Telegram details

4. Run the Pi app

   python3 main.py

Backend setup, Telegram bot setup, and full deployment are documented in SETUP_GUIDE.md.

📦 Project Structure

fingerprint-camera-door-lock/
├── Rasberry pi/
│   ├── main.py                   # Main Pi flow (PIN → Face → Fingerprint)
│   ├── raw_data.py               # Raw sensor reads logger (week 1a)
│   ├── face_recognition_folder/  # Face recognition modules
│   ├── fingerprint/              # R307 fingerprint interface
│   ├── numpad/                   # Keypad input
│   └── servo/                    # Door lock control
├── Backend/
│   ├── server.py                 # Flask API (VPS)
│   └── requirements.txt
├── supabase/
│   └── setup.sql                 # Database schema
├── assets/
│   ├── banner.png
│   ├── gpio.png
│   └── raspberrypi5.jpg
├── README.md
└── LICENSE

🌐 API Endpoints (Backend)

Base URL: <BACKEND_URL> (for example: https://your-domain.example.com/doorlock)

• GET /health — Health check
• POST /api/verify-user — Verify user credentials
• POST /api/log-access — Log access attempt
• GET /api/get-users — List users
• POST /api/add-user — Add user
• POST /api/upload-face-image — Upload face image
• POST /api/retrain-model — Retrain face model
• GET /api/get-model-info — Get current model info
• GET /models/trained_model.pkl — Download current model
• GET /api/get-access-logs — Access logs
• POST /api/sync-status — Pi status sync

🧪 Raw Sensor Reads (Week 1a)

Use Rasberry pi/raw_data.py to record unprocessed sensor logs for PIN keypad, fingerprint, and camera recognition. The script prints timestamped logs suitable for screen recording.

🔐 Security Best Practices

• Never commit .env or secrets
• Use HTTPS for backend
• Enable Supabase RLS policies
• Rotate keys periodically
• Monitor access logs

📄 License

This project is licensed under the Boost Software License 1.0 — see the LICENSE file for details.

👤 Author

Hardik Garg (@sasukeuchiha14)

🙏 Acknowledgments

• Face Recognition library (ageitgey)
• Adafruit Fingerprint Sensor resources
• Supabase team
• Raspberry Pi Foundation