Face Recognition Attendance System

A comprehensive Python-based face recognition attendance system that automatically detects and recognizes faces from a webcam feed and marks attendance in CSV files. The system uses OpenCV for computer vision tasks and machine learning algorithms for face recognition.

Features

• Real-time Face Detection: Uses Haar Cascade classifiers for accurate face detection
• Face Recognition: Implements LBPH (Local Binary Pattern Histogram) algorithm for face recognition
• Automatic Attendance Marking: Generates timestamped CSV attendance files
• User-friendly Interface: Menu-driven console application
• Student Database Management: Maintains student details in CSV format
• Camera Testing: Built-in camera functionality testing
• Training System: Automated face model training from captured images

System Requirements

• Python 3.7 or higher
• Webcam/Camera device
• Windows/Linux/macOS operating system
• Minimum 4GB RAM recommended

Installation Guide

Step 1: Clone or Download the Project

git clone <repository-url>
cd Face-Recognition-Attendance-System

Step 2: Install Python Dependencies

pip install -r requirements.txt

Note: If you encounter issues with dlib installation on Windows, you may need to install Visual Studio Build Tools or use pre-compiled wheels.

Step 3: Verify Installation

Run the camera check to ensure your webcam is working:

python check_camera.py

Detailed Usage Instructions

Step 1: Initial Setup

1. Ensure all project files are in the same directory
2. Verify that haarcascade_frontalface_default.xml and haarcascade_eye.xml files are present
3. Create necessary directories if they don't exist:
   • TrainingImage/ - for storing captured face images
   • TrainingImageLabel/ - for storing trained model files
   • StudentDetails/ - for storing student information
   • Attendance/ - for storing attendance records

Step 2: Run the Main Application

python main.py

Step 3: Using the Menu System

Option 1: Check Camera

• Tests your webcam functionality
• Displays live video feed with face detection rectangles
• Press 'q' to quit the camera test
• Useful for verifying camera setup before proceeding

Option 2: Capture Faces

• Purpose: Capture training images for new students
• Process:
  1. Enter Student ID (numeric only)
  2. Enter Student Name (alphabetic only)
  3. Position your face in front of the camera
  4. The system will automatically capture 100+ images
  5. Images are saved as Name.ID.ImageNumber.jpg in TrainingImage/ folder
  6. Student details are automatically added to StudentDetails.csv
• Tips:
  • Ensure good lighting conditions
  • Move your head slightly for different angles
  • Keep face clearly visible and centered
  • Press 'q' to stop capturing early if needed

Option 3: Train Images

• Purpose: Create the face recognition model from captured images
• Process:
  1. Automatically processes all images in TrainingImage/ folder
  2. Extracts facial features using LBPH algorithm
  3. Creates Trainner.yml file in TrainingImageLabel/ folder
  4. Shows progress counter during training
• Requirements: Must have captured images using Option 2 first
• Duration: Depends on number of images (typically 30-60 seconds)

Option 4: Recognize & Attendance

• Purpose: Start the attendance marking system
• Process:
  1. Loads the trained model (Trainner.yml)
  2. Starts webcam feed for real-time recognition
  3. Detects and recognizes faces in the video stream
  4. Displays confidence percentage for each recognition
  5. Marks attendance for recognized faces (confidence > 70%)
  6. Press 'q' to stop and save attendance
• Output: Creates timestamped CSV file in Attendance/ folder
• File Format: Attendance_YYYY-MM-DD_HH-MM-SS.csv

Option 5: Quit

• Safely exits the application

Step 4: Understanding the Output Files

Student Details (StudentDetails/StudentDetails.csv)

Id,Name
1,John
2,Alice
3,Bob

Attendance Records (Attendance/Attendance_YYYY-MM-DD_HH-MM-SS.csv)

Id,Name,Date,Time
1,John,2024-01-15,09:30:25
2,Alice,2024-01-15,09:31:10

File Structure

Face-Recognition-Attendance-System/
├── main.py                          # Main application entry point
├── Capture_Image.py                 # Face capture functionality
├── Train_Image.py                   # Model training functionality
├── Recognize.py                     # Face recognition and attendance
├── check_camera.py                  # Camera testing utility
├── face_recog.py                    # Alternative recognition script
├── automail.py                      # Email functionality (future feature)
├── requirements.txt                 # Python dependencies
├── haarcascade_frontalface_default.xml  # Face detection model
├── haarcascade_eye.xml             # Eye detection model
├── TrainingImage/                   # Captured face images
├── TrainingImageLabel/              # Trained model files
├── StudentDetails/                  # Student information
├── Attendance/                      # Attendance records
└── README.md                        # This file

Troubleshooting

Common Issues and Solutions

1. Camera Not Working

   • Check if camera is connected and not used by other applications
   • Try changing camera index in code (0, 1, 2, etc.)
   • Ensure camera permissions are granted

2. Face Not Detected

   • Ensure adequate lighting
   • Position face clearly in camera view
   • Check if haarcascade_frontalface_default.xml file exists

3. Low Recognition Accuracy

   • Capture more training images (100+ recommended)
   • Ensure diverse angles and lighting during capture
   • Retrain the model with new images

4. Import Errors

   • Verify all dependencies are installed: pip install -r requirements.txt
   • For dlib issues on Windows, install Visual Studio Build Tools

5. File Permission Errors

   • Run with administrator privileges if needed
   • Check write permissions for output directories

Performance Optimization

• For Better Accuracy: Capture 150-200 images per person
• For Faster Processing: Reduce camera resolution in code
• For Multiple People: Ensure distinct facial features in training data

Technical Details

Algorithms Used

• Face Detection: Haar Cascade Classifiers
• Face Recognition: LBPH (Local Binary Pattern Histogram)
• Image Processing: OpenCV library

Recognition Confidence

• High Confidence: > 70% (marked as present)
• Medium Confidence: 30-70% (displayed but not marked)
• Low Confidence: < 30% (marked as unknown)

Future Enhancements

• Google Sheets integration for cloud-based attendance
• Email notifications for attendance reports
• Web-based interface
• Mobile app integration
• Advanced deep learning models
• Multi-camera support
• Real-time dashboard

Contributing

1. Fork the repository
2. Create a feature branch
3. Make your changes
4. Test thoroughly
5. Submit a pull request

Support

For issues and questions:

1. Check the troubleshooting section above
2. Review the code comments for technical details
3. Create an issue in the repository

---