🚨 Real-Time Intruder Detection and Notification System

A smart security system using Computer Vision, face recognition and real-time notifications to detect unknown intruders and alert the homeowner via SMS and a loud audio beep.

---

📌 Overview

In today's world, ensuring secure residential spaces is crucial. Traditional systems like fiber sensors or accelerometers are often unreliable, expensive, and environmentally sensitive. Our system addresses these limitations by using machine learning, face recognition, and IoT-enabled communication to detect intruders in real time and immediately notify the owner.

This system captures faces via webcam, checks if they are known, and if not:

• Triggers a beep alert.
• Sends an SMS notification to the homeowner.

---

🧰 Technologies Used

Tool/Library	Purpose
Python	Core programming language
OpenCV	Real-time video processing
face_recognition	Face detection and encoding
Twilio API	Sending SMS alerts
winsound	Audio alerts (Windows only)
Webcam + Speaker	Real-time monitoring and beeping

---

📁 Project Structure

📦 Real-Time-Intruder-Detection/
 ┣ 📂 known_faces/         # Store known face images here
 ┣ 📜 intruder_alert.py    # Main Python script
 ┗ 📜 README.md            # This file

---

🚀 How to Run

1. Clone the repository

git clone https://github.com/your-username/Real-Time-Intruder-Detection.git
cd Real-Time-Intruder-Detection

2. Install dependencies

pip install opencv-python face_recognition twilio

3. Add known faces

Place clear frontal face images (JPG or PNG) of known people in the known_faces/ folder.
The file names (excluding the extension) will be used as the person's name.

known_faces/
├── Alice.jpg
├── Bob.png

4. Update credentials

In intruder_alert.py, update these values:

account_sid = 'YOUR_ACCOUNT_SID'
auth_token = 'YOUR_AUTH_TOKEN'
twilio_phone_number = '+1234567890'
your_phone_number = '+0987654321'
known_faces_dir = "known_faces"

5. Run the system

python intruder_alert.py

Press q to quit the video stream.

---

⚙️ How It Works

1. Loads all known faces from the known_faces folder and encodes them.
2. Captures real-time video using your webcam.
3. For each detected face:
   • If it matches a known encoding, the name is displayed.
   • If it's unknown, a beep alert is played and an SMS is sent via Twilio.

---

🔐 Benefits

• No need for traditional sensors (e.g., accelerometers)
• Resilient to environmental factors like temperature
• Real-time alert system
• Works with basic hardware (camera + speaker)

---

💡 Future Enhancements

• Add GUI for registering new faces
• Email and mobile app notifications
• Deploy on Raspberry Pi for portability
• Add face detection logs and timestamps
• Multi-camera support
• Face dataset sync via cloud

---

🙌 Credits

Built with:

• OpenCV
• face_recognition
• Twilio API
• Python

---

🛑 Disclaimer

This project is for educational and ethical use only. Do not use for unauthorized surveillance.

---

📞 Contact

For queries, improvements, or collaborations, feel free to reach out via GitHub or LinkedIn