Real-time fire, smoke, motion, and face detection system with a beautiful web dashboard, Telegram alerts, and optional Raspberry Pi hardware support (camera, MQ2 gas sensor, audio alarm).

🔐 Security First: Telegram credentials live in a .env file (never committed). settings.json only contains non-sensitive options.

• 🔥 Fire Detection - YOLO11 model trained for fire detection
• 💨 Smoke Detection - Visual smoke detection with YOLO11
• 👤 Face Detection - MediaPipe-powered face detection and tracking
• 🏃 Motion Detection - OpenCV-based motion tracking
• 🌐 Web Dashboard - Beautiful, modern real-time dashboard
• 📱 Telegram Alerts - Instant notifications with images
• 🔊 Text-to-Speech - Browser-based audio playback using gTTS
• 🎯 Bounding Box Toggle - Show/hide detection overlays
• 📦 Easy Setup - Clone and run in minutes

• Python 3.10 or higher
• Webcam or IP camera (MJPEG/RTSP stream)
• Chromium-based browser (Chrome, Edge, Brave) - Recommended for best compatibility
  • ⚠️ Firefox may have issues with video streaming and audio playback

# 1. Clone the repository
git clone https://github.com/OliseNS/Inferno.git
cd Inferno

# 2. Create virtual environment
python -m venv .venv
source .venv/bin/activate  # On Windows: .venv\Scripts\activate

# 3. Install dependencies
pip install -r requirements.txt

# 4. Set up environment variables (for Telegram - optional)
cp .env.example .env
# Edit .env with your favorite text editor:
nano .env  # or vim, code, etc.
# Add your TELEGRAM_BOT_TOKEN and TELEGRAM_CHAT_ID

# Run the full system (web dashboard + detection)
python webserver.py

Open your Chromium-based browser and navigate to:

• Local: http://localhost:8080
• Network: http://<your-ip>:8080

Run detection only without the web interface:

python detection_system.py

system.model_path in settings.json points at the YOLO export (default: models/fire_smoke_yolo11_ncnn_model). NCNN exports must live in a directory whose name ends with _ncnn_model. Override with another folder or .pt path.

The dashboard can append/update these in .env when you save token/chat ID (non-empty fields only).

• OS: Linux, macOS, or Windows 10/11
• Python: 3.10 or higher
• RAM: 2GB minimum (4GB recommended)
• Storage: 500MB for dependencies + model files
• Camera: USB webcam or IP camera (MJPEG/RTSP stream)

• Model: Raspberry Pi 4 (4GB RAM or higher)
• OS: Raspberry Pi OS (64-bit recommended)
• Camera: Pi Camera Module or USB webcam
• Optional: MQ2 gas sensor, audio speaker for alarm

On first run, settings.json is automatically created with default values. You can modify it directly or use the web dashboard.

Option 1: Built-in Camera (default)

{
  "system": {
    "camera_url": "",
    "camera_index": 0
  }
}

Option 2: IP Camera Stream

{
  "system": {
    "camera_url": "http://192.168.1.100:8081/video"
  }
}

Option 3: Camera by Index

{
  "system": {
    "camera_url": "1"
  }
}

1. Create a bot with @BotFather
2. Get your chat ID from @userinfobot
3. Add to .env:

TELEGRAM_BOT_TOKEN=your_bot_token_here
TELEGRAM_CHAT_ID=your_chat_id_here

4. Enable in dashboard or settings.json:

{
  "telegram": {
    "enabled": true,
    "cooldown": 30
  }
}

• Toggle fire, smoke, motion, and face detection on/off
• Show/hide bounding boxes on video feed
• Adjust detection sensitivity in settings

• Live camera feed with detection overlays
• System status indicator (Normal/Fire/Smoke/Motion/Face)
• Frame counter and system uptime
• Recent face gallery
• Detection image gallery

• Type any message and hear it through your browser
• TTS history with click-to-replay
• Uses Google Text-to-Speech (gTTS)

• Test connection with one click
• View credentials status
• Adjust notification cooldown
• Automatic alerts with detection images

Problem: Black screen or "Waiting for camera..."

Solutions:

1. Check camera connection: ls /dev/video* (Linux)
2. Try different camera index in settings (0, 1, 2, etc.)
3. Ensure no other application is using the camera
4. For IP cameras, verify the stream URL is accessible

Problem: No sound when using TTS

Solutions:

1. Use a Chromium-based browser (Chrome, Edge, Brave)
2. Check browser audio permissions
3. Ensure speakers/headphones are connected and volume is up
4. Try clicking the play button in the browser if autoplay is blocked

Problem: Video stuttering or TTS not playing

Solution: Firefox has known issues with MJPEG streams and audio playback. Please use Chrome, Edge, or Brave for the best experience.

Problem: No detections appearing

Solutions:

1. Check if detection types are enabled in dashboard
2. Verify model path in settings.json points to correct YOLO model
3. Ensure proper lighting for fire/smoke detection
4. Check console logs for errors: tail -f logs/inferno.log

Problem: Test message fails

Solutions:

1. Verify token and chat ID in .env are correct
2. Check bot is not blocked
3. Ensure internet connection is stable
4. Check Telegram is enabled in settings
5. Verify cooldown period hasn't been triggered

Problem: Slow frame rate or system lag

Solutions:

1. Reduce video resolution in camera settings
2. Increase detection_interval in settings (e.g., 1.0 seconds)
3. Disable face detection if not needed (most CPU-intensive)
4. Use NCNN model for better performance
5. Ensure Pi has adequate cooling

Problem: "Permission denied" when accessing camera or GPIO

Solutions:

# Add user to video group (Linux)
sudo usermod -a -G video $USER

# For GPIO on Raspberry Pi
sudo usermod -a -G gpio $USER

# Reboot after changes
sudo reboot

All console output (stdout/stderr) is automatically logged to logs/inferno.log. The logs/ directory is created automatically.

# View live logs
tail -f logs/inferno.log

# Search for errors
grep "ERROR" logs/inferno.log

Contributions are welcome! Please feel free to submit a Pull Request. For major changes, please open an issue first to discuss what you would like to change.

1. Fork the repository
2. Create your feature branch (git checkout -b feature/AmazingFeature)
3. Commit your changes (git commit -m 'Add some AmazingFeature')
4. Push to the branch (git push origin feature/AmazingFeature)
5. Open a Pull Request

• Ensure good lighting conditions
• Position camera to have clear view of monitored areas
• Test with different angles for optimal detection

• Adequate lighting is crucial for face detection
• Faces should be reasonably front-facing
• Works best at 1-3 meters distance

• Minimize background movement (curtains, fans)
• Adjust min_area in code if getting too many false positives
• Works better with static camera mount

• Start with fewer detection types enabled
• Gradually enable more as needed
• Monitor system resources (CPU, RAM)
• On Raspberry Pi, consider using lower resolution camera

• Modern UI: Beautiful gradient design with smooth animations
• Real-time Processing: Low-latency detection with optimized frame processing
• Flexible Configuration: Everything configurable via JSON or web UI
• Privacy-First: All processing happens locally, no cloud dependencies
• Production Ready: Logging, error handling, and graceful degradation
• Extensible: Easy to add new detection types or notification channels

If you encounter any issues or have questions:

1. Check the Troubleshooting section
2. Review logs in logs/inferno.log
3. Open an issue on GitHub

OliseNS — GitHub

Built with ❤️ using Python, Flask, OpenCV, YOLO11, and MediaPipe.

No license file is included yet; contact the author for usage terms.

# Start system
python webserver.py

# Check logs
tail -f logs/inferno.log

# Update dependencies
pip install -r requirements.txt --upgrade

# Reset configuration
rm settings.json  # Will be recreated with defaults

# Clear old detections
rm -rf faces/* detections/* voices/*

⭐ If you find this project useful, please consider giving it a star on GitHub!