Face Recognition Attendance System

A Python-based GUI desktop application that uses real-time face recognition to automatically mark employee attendance. Built with OpenCV, face_recognition, and Tkinter, it records attendance data to an Excel sheet with attendance percentage calculations.

• Face Recognition Login — Identifies registered employees via webcam and marks attendance instantly.
• New User Registration — Capture a face photo via webcam and store employee details.
• Duplicate Prevention — Prevents marking attendance more than once per day for the same employee.
• Attendance Percentage — Automatically calculates attendance percentage based on working days (excluding Sundays and Indian public holidays).
• Excel-Based Storage — All employee and attendance data is persisted in .xlsx files — no database required.

PythonProject/
│
├── app.py                  # Main application entry point
├── employees.xlsx          # Stores registered employee records (ID, Name)
├── attendance.xlsx         # Stores attendance logs (Date, Time, %, etc.)
├── faces/                  # Directory storing registered face images
│   └── <empID>_<name>.jpg
└── README.md

• Python 3.8+

Install all dependencies using pip:

pip install opencv-python face_recognition numpy pandas openpyxl holidays

Note: face_recognition requires dlib. On Windows, it is recommended to install a prebuilt wheel or use CMake + Visual Studio Build Tools.

pip install cmake
pip install dlib
pip install face_recognition

• A working webcam (built-in or external).

1. Clone the repository

   git clone <your-repo-url>
   cd PythonProject

2. Install dependencies

   pip install opencv-python face_recognition numpy pandas openpyxl holidays

3. Run the application

   python app.py

1. Click "New User Registration" on the main window.
2. Enter the Employee ID and Name.
3. Click "Register and Capture Face".
4. A webcam window will open — look at the camera.
5. Press SPACE to capture your face or ESC to cancel.
6. The employee is saved to employees.xlsx and their face image to the faces/ folder.

1. Click "Login (Face Recognition)" on the main window.
2. The webcam activates and scans for registered faces.
3. If a match is found:
   • A green rectangle appears around the face.
   • Attendance is marked in attendance.xlsx with the current date, time, and attendance percentage.
   • A success dialog is shown.
4. If the face is unrecognized, a red rectangle with "Unknown" is displayed.
5. Press ESC to close the webcam window manually.

The attendance.xlsx file tracks the following fields:

Working days exclude Sundays and Indian public holidays (via the holidays library).

These can be changed at the top of app.py.

• Only one face is processed per recognition session (first match wins).
• Attendance percentage is calculated from January 1st of the current year.
• Face recognition accuracy may decrease in low-light conditions.
• face_recognition library can be slow on machines without a GPU.

This project is created for academic purposes as part of a 6th Semester Python Lab.