FALCON is an AI-powered real-time surveillance system that detects weapons using YOLOv8 and triggers instant alerts.
It integrates computer vision, backend systems, and a web dashboard to monitor, log, and visualize incidents.
- Real-time weapon detection using YOLOv8
- Live camera / RTSP stream processing
- Instant alert system (Twilio SMS)
- Interactive dashboard with analytics
- Detection reports with images & videos
- User authentication (Login/Register)
- Edge AI support (Raspberry Pi)
- Camera captures live video (RTSP / USB camera)
- Frames are processed using YOLOv8
- If a weapon is detected:
- Image/video is saved
- Detection data is stored in database
- Alert is sent via Twilio
- Dashboard updates with real-time data
- Model: YOLOv8 (Custom Trained)
- mAP@0.5: ~81%
- Precision: ~85%
- Recall: ~75%
- Epochs: ~50
- FPS: ~10
- Dataset Size: ~9,633 Images
| Device | Inference Time |
|---|---|
| GPU (Laptop) | 40–60 ms |
| CPU | 120–200 ms |
| Raspberry Pi | 400–800 ms |
- GPU: 15–25 FPS
- CPU: 5–10 FPS
- Raspberry Pi: 1–3 FPS
- Backend: Flask (Python)
- AI Model: YOLOv8
- Database: SQLite
- Frontend: HTML, CSS, JavaScript
- Alerts: Twilio API
- Computer Vision: OpenCV
git clone https://github.com/Astik97/FALCON.git
cd FALCON
python -m venv venv source venv/bin/activate # Linux/Mac venv\Scripts\activate # Windows
pip install -r requirements.txt
Create .env file:
TWILIO_SID=your_sid TWILIO_AUTH=your_auth TWILIO_PHONE=your_number
python app.py Open in browser:http://127.0.0.1:5000
.envfile is excluded- API keys are protected
- Sensitive data not stored in repo
- Performance drops in low-light conditions
- False positives may occur
- Raspberry Pi has limited processing power
- Requires stable network for RTSP streams
- Multi-camera support
- Cloud deployment (AWS / Azure)
- Model optimization (TensorRT / ONNX)
- Facial recognition integration
- Advanced alert system (email + app notifications)
Astik Mohapatra
B.Tech Computer Science Engineering
Government College of Engineering Keonjhar (CGPA: 8.1/10, 2026)
astikm7007@gmail.com
https://linkedin.com/in/astik-mohapatra
https://github.com/Astik97
Give it a ⭐ on GitHub and share!