DetectCV: Real-Time Face Recognition Attendance System

DetectCV is a Flask-based web application that provides real-time face recognition for attendance tracking. It supports multi-camera selection, adjustable recognition tolerance, and a clean dashboard for managing known faces and attendance logs.

Features

• Real-Time Face Recognition: Uses face-recognition and OpenCV to identify faces from webcam input.
• Multi-Camera Support: Choose between multiple camera sources (e.g., built-in, USB) via dropdown.
• Dynamic Tolerance: Adjust the face distance threshold to control matching strictness.
• Web-Based Dashboard: Responsive UI built with Bootstrap 5 for uploading reference images and viewing live feed.
• Attendance Management:
  • Automatic logging of first-time recognized faces to Attendance.csv with timestamp.
  • View attendance records through a web interface.
  • Download attendance as CSV and clear logs.

Project Structure

DetectCV/
├── app.py
├── config.py
├── requirements.txt
├── known_faces/           # Reference images (.jpg, .png)
├── Attendance.csv         # Generated attendance log
├── static/
│   ├── css/
│   │   └── styles.css     # Custom CSS
│   └── js/
│       └── script.js      # Camera & tolerance controls
└── templates/
    ├── index.html         # Main dashboard
    └── attendance.html    # Attendance log view

Dependencies

The following key libraries and tools are used in this project:

• Flask (2.3.2): Web framework for routes, templating, and HTTP handling.
• OpenCV (4.5.4.60): Video capture and image processing.
• face-recognition (1.3.0): Face encoding and matching built atop dlib.
• dlib (19.24.2): Underlying C++ library for facial landmark detection and embedding.
• NumPy (1.19.5): Numerical operations, especially on image arrays.
• CMake (3.18.4): Build tool required by dlib installation.

Architecture

DetectCV follows a modular MVC-like pattern:

• Configuration (config.py): Centralizes settings such as tolerance threshold, camera index, and file paths.
• Model (Face Recognition):
  • Face Encoding Loader: Reads images from known_faces/, computes 128‑d embeddings via face-recognition.
  • Distance Matching: Compares live-frame embeddings against known encodings using Euclidean distance.
• View (Templates):
  • Dashboard (index.html): Bootstrap-based UI for camera/tolerance controls and live feed.
  • Attendance Log (attendance.html): Tabular listing with download/clear actions.
• Controller (app.py):
  • Routes: Handle HTTP GET/POST for dashboard, video stream, attendance viewing, file upload.
  • Video Generator: Captures frames, detects faces, matches names, logs attendance, streams MJPEG.
• Static Files:
  • CSS/JS: Custom scripts for dynamic slider/camera selection and styling.

Installation

1. Clone the repository:

   git clone https://github.com/yourusername/DetectCV.git
   cd DetectCV

2. Create a Python virtual environment (optional but recommended):

   python -m venv .venv
   source .venv/bin/activate  # Linux/Mac
   .venv\Scripts\activate    # Windows

3. Install dependencies:

   pip install -r requirements.txt

4. Prepare reference images:

   • Add .jpg, .jpeg, or .png files named after the person (e.g., alice.jpg) into known_faces/.

Configuration

Edit config.py to adjust:

• TOLERANCE: Face distance threshold (default 0.5).
• DEFAULT_CAMERA: Camera index (default 0).
• REFERENCE_FOLDER: Path to known faces directory.
• ATTENDANCE_FILE: CSV output for attendance logs.

Usage

1. Run the application:

   python app.py

2. Access the dashboard: Open a web browser and navigate to http://127.0.0.1:5000/.

3. Upload reference images:

   • Use the file upload form in the navbar to add new faces.
   • The app will automatically reload known faces.

4. View live recognition:

   • Select camera source and adjust tolerance in the Settings panel.
   • The Live Feed panel shows real-time face detection.

5. Manage attendance:

   • Click Log in the navbar to view all recognized attendance records.
   • Download the CSV or clear the log using the provided buttons.

---