Real-Time Face Recognition System

This project is a real-time face recognition system built using Python, OpenCV, and the face_recognition library. It allows you to recognize known faces via webcam, log each recognition with a timestamp, and manage face data through a simple GUI.

⚠️ Note: This project does not include any personal images, only few Unsplash images for comparison. You are expected to add your own face data following the instructions below.

---

Features

• Real-time face recognition using webcam
• Add new faces via webcam or file upload
• Timestamp logging of recognized people in log.csv
• Simple GUI built with Tkinter
• Clean architecture and modular code

---

Requirements

Install dependencies using:

pip install -r requirements.txt

Folder Structure

real_time_face_recognition/ ├── images/ # Stores folders for each person (e.g., images/John/) ├── encodings.pickle # Pickle file with facial encodings ├── log.csv # Logs of recognized faces with timestamps ├── encode_faces.py # Encodes faces into pickle file ├── add_face.py # Script to add new faces ├── gui_app.py # Main GUI application ├── recognizer.py # CLI face recognition (optional) ├── requirements.txt └── README.md

How to use

Clone repo:

git clone https://github.com/M26I/real_time_face_recognition
cd real_time_face_recognition

Add new Face:

python add_face.py

You will be prompted to:

Enter the name of the person (e.g., John)

Choose between:

'webcam' — take a photo using your webcam

'file' — provide a path to an image on your computer

Captured or uploaded images will be stored under images/{name}/.

The script will then automatically update facial encodings.

Run the app:

python gui_app.py

The webcam will open, and known faces will be recognized in real-time. Every recognition event is saved to log.csv with the name and timestamp.

Notes

• You can delete all face data by removing the contents of the images/ folder and deleting encodings.pickle.

• Never commit personal images to a public repository.

• Make sure each face you add has good lighting and a clear frontal view for better accuracy.

Author

M26I

---

© 2025 M26I – For educational/portfolio use only.
Unauthorized use or redistribution without credit is prohibited.