Face Recognition Attendance System

📋 Overview

The Face Recognition Attendance System is a real-time, privacy-focused solution for automated attendance tracking using facial recognition. It leverages computer vision and machine learning to detect and recognize faces from a webcam, logging attendance with timestamps in a secure and efficient manner.

• No images are stored—only face encodings are kept for privacy and speed.
• Unknown faces are also logged (once per day) for audit purposes.
• Robust error handling and logging ensure reliability in real-world use.

---

✨ Key Features

• Real-time face detection and recognition
• Automated attendance logging with timestamps
• User-friendly interface with visual feedback overlays
• Multi-face detection capability
• Privacy-focused: only encodings, not images, are stored
• Configurable system parameters via config.json
• Comprehensive error logging to attendance_system.log
• Logs both known and unknown faces (once per day)

---

🚀 Quick Start

Prerequisites

• Python 3.6+
• Visual Studio Build Tools with C++ (for dlib/face_recognition)
• CMake
• Webcam (720p minimum)
• 4GB+ RAM

Installation

1. Install Visual Studio Build Tools

   • Download Visual Studio Community Edition
   • Select "Desktop Development with C++"
   • Complete installation and restart

2. Clone the repository

   git clone https://github.com/lifaet/Face-Recognition-Attendance-System.git
   cd Face-Recognition-Attendance-System

3. Install Python dependencies

   pip install cmake
   pip install dlib
   pip install face-recognition
   pip install numpy
   pip install opencv-python

---

📂 Project Structure

Face-Recognition-Attendance-System/
├── fras.py                 # Main application
├── add_attendees.py        # Register new attendees (encodings only)
├── encodings.json          # Stores face encodings and names
├── attendance.csv          # Attendance records
├── config.json             # System configuration
├── attendance_system.log   # System logs
└── README.md               # Documentation

---

⚙️ Configuration

The system is configured via config.json:

{
    "path": "attendees",
    "frame_skip": 2,
    "face_recognition_threshold": 0.50,
    "attendance_file": "attendance.csv",
    "ui": {
        "analyzing_text": "Analyzing...",
        "welcome_text": "Welcome,",
        "unknown_text": "Unknown Person",
        "display_time": 3
    }
}

• frame_skip: Process every Nth frame for speed.
• face_recognition_threshold: Lower is stricter (default 0.50).
• display_time: Seconds to show welcome message.

---

📸 Registering Attendees

Use add_attendees.py to add new people to the system. This script captures a face from the webcam, extracts its encoding, and saves it to encodings.json.

Usage

1. Run the script:

   python add_attendees.py

2. Enter the attendee's name.
3. Position their face in the webcam and press c to capture.
4. The encoding is saved to encodings.json.

Note: No images are stored—only the encoding is kept for privacy.

---

🖥️ Running the Attendance System

1. Start the application:

   python fras.py

2. The webcam opens and waits for a face.
3. When a face is detected:
   • Shows "Analyzing..." for 1–2 seconds.
   • If recognized, shows "Welcome, NAME!" and logs attendance.
   • If not recognized, shows "Unknown Person" and logs "Unknown" attendance.
4. Press q or Ctrl+C to exit.

---

🔍 Technical Details

Face Recognition Process

1. Face Detection

   • Locates faces in video feed
   • Processes at 1/4 resolution for performance

2. Face Recognition

   • Converts detected faces to encodings
   • Matches against known face encodings from encodings.json
   • Threshold-based verification

3. Attendance Marking

   • Automatic date and time stamping
   • Duplicate entry prevention (one entry per person per day)
   • CSV format storage

UI Features

• Status messages for:
  • Face analysis in progress
  • Welcome messages
  • Unknown person alerts
• Visual indicators:
  • Green rectangle around detected faces
  • Checkmark for successful recognition
• Semi-transparent overlay

---

📊 Performance

• Frame skipping for optimal performance (frame_skip in config)
• Configurable recognition threshold
• Efficient image processing and memory usage

---

🔧 Error Handling & Logging

• All errors and events are logged to attendance_system.log.
• Graceful shutdown on errors or Ctrl+C.
• Frame-level errors are logged and skipped without crashing the system.

---

📝 Attendance Logging

• Attendance is logged in attendance.csv as:

  Name,Date,Time
  

• Each person (including "Unknown") is logged only once per day.

---

🎯 Applications

Educational

• Classroom attendance
• Event tracking
• Library access

Corporate

• Employee attendance
• Meeting participation
• Visitor tracking

Events

• Participant check-in
• Access control
• Session tracking

---

🔧 Troubleshooting

Common Issues

1. Camera not detected

   • Check camera connections
   • Verify camera permissions

2. Recognition issues

   • Ensure good lighting
   • Update reference encodings
   • Adjust recognition threshold

3. Performance issues

   • Increase frame skip value
   • Check system resources
   • Update hardware drivers

---

🔒 Security & Privacy

• Only face encodings (not images) are stored.
• No images are saved after registration.
• Attendance logs do not include images or biometric data, only names and timestamps.

---

📜 License

This project is for educational and personal use.