Face Recognition Attendance System

This project is a Face Recognition Attendance System developed using Python, OpenCV, and face_recognition. The system recognizes faces through your webcam and marks attendance for registered users. It saves the attendance data in a CSV file with timestamps and provides a GUI interface for user interaction.

Features

• User Registration: Register a new user by uploading their photo.
• Face Recognition: The system captures faces in real-time using a webcam and matches them with registered users.
• Attendance Logging: Attendance is logged with a timestamp whenever a registered user's face is recognized.
• View Attendance: View the recorded attendance in a CSV format.
• Download Attendance: Download the attendance data as a CSV file.
• Start/Stop Recognition: Start and stop the real-time face recognition process with buttons.

Requirements

• Python 3.x
• Libraries:
  • opencv-python
  • face_recognition
  • numpy
  • tkinter
  • os
  • datetime
  • csv
  • logging
  • platform
  • subprocess
  • PIL (for image file handling)

You can install the necessary libraries using pip:

pip install opencv-python face_recognition numpy Pillow

Setup

1. Create known_faces Directory:
   Create a directory named known_faces in the project root directory. This is where the photos of registered users will be stored.

2. Register New Users:

   • Enter the user name in the provided text field.
   • Select an image file (JPEG/PNG) containing the user's face when prompted.
   • The image will be saved in a folder named after the user inside the known_faces directory.

3. Run the System:

   • Press the Start Attendance button to begin the face recognition process.
   • The system will continuously monitor for faces, and when a registered user's face is detected, their attendance will be recorded with a timestamp.
   • To stop the process, press the Stop Attendance button.

4. View Attendance:

   • Click on View Attendance to open a window that displays the recorded attendance.

5. Download Attendance:

   • Click on Download Attendance to download the attendance data as a CSV file to your local machine.

Directory Structure

Face_Recognition_Attendance_System/
├── known_faces/            # Folder containing user photos
│   └── JohnDoe/            # Subfolder for each registered user
│       └── john_doe.jpg    # User's photo for recognition
├── attendance.csv          # File to store attendance data
├── attendance_system.py    # Main Python script (this file)
└── attendance_system.log   # Log file to track system activity

How It Works

1. User Registration: When a user registers, their face encoding is extracted and stored in the known_faces directory. Each user has their own folder, and photos are saved inside it.
2. Face Recognition: The system uses the webcam to capture frames, performs face detection, and compares the detected faces with the stored face encodings. When a match is found, the user's attendance is logged in the attendance.csv file.
3. Attendance: The system records the user's name and timestamp in the CSV file every time they are recognized.

Dependencies

• opencv-python: Used for capturing video frames from the webcam.
• face_recognition: Used for face detection and encoding.
• numpy: Required by face_recognition for array manipulations.
• tkinter: Used for building the graphical user interface (GUI).
• Pillow: For handling images.

Troubleshooting

• Unable to access camera: Make sure your webcam is connected and not being used by another application. On macOS, you may need to adjust camera permissions.
• No face recognized: Ensure the image is clear and the person's face is fully visible. The system may require good lighting and a clear view of the face to work effectively.

License

This project is licensed under the MIT License - see the LICENSE file for details.

Acknowledgments

• face_recognition: The face_recognition library provides a simple API for face detection and recognition.
• OpenCV: The opencv-python library helps to capture frames from the webcam and process the video.