Real-Time Face Recognition and Attendance System

This project implements a real-time face recognition and attendance tracking system using OpenCV, Firebase, and the face_recognition library. The system captures video frames from a webcam, detects and recognizes faces, and uploads relevant entry and exit data to Firebase Realtime Database and Storage.

Table of Contents

• Project Description
• Features
• Installation
• Usage
• File Structure
• Project Details
  • Firebase Setup
  • Face Detection and Recognition
  • Handling and Storing Data
• Contributing
• License

Project Description

The Real-Time Face Recognition and Attendance System uses computer vision techniques to identify and track the entry and exit of individuals from a video feed. It detects faces using Haar cascades and recognizes them with pre-stored encodings. This information is recorded in Firebase, including entry and exit images, timestamps, and attendance status.

Features

• Real-time face detection and recognition using OpenCV and the face_recognition library.
• Attendance tracking with entry and exit times, images, and status classification.
• Data storage using Firebase Realtime Database and Firebase Storage.
• Threaded processing for encoding faces concurrently to improve performance.

Installation

1. Clone the repository:

   git clone https://github.com/your-username/real-time-face-recognition-attendance.git
   cd real-time-face-recognition-attendance

2. Install dependencies:

   pip install opencv-python cvzone firebase-admin face_recognition numpy

3. Set up Firebase credentials:

   • Place your serviceAccKey.json (downloaded from Firebase) in the root folder.
   • Update the databaseURL and storageBucket values in the code to match your Firebase project's settings.

4. Add face data:

   • Create and store your face encodings in the EncodeFile.p using the AddDataToData.py script or similar. Ensure that the peopleFaceList and peopleID lists are correctly formatted.

Usage

1. Run the main script:

   python main_script.py

2. The webcam feed will open and start detecting faces. If a known face is detected, the system will log the entry and exit data to Firebase.

File Structure

📦project-root
 ┣ 📜main_script.py       # Main script for running face recognition
 ┣ 📜AddDataToData.py     # Script to add and encode face data
 ┣ 📜EncodeFile.p         # Encoded face data
 ┣ 📜serviceAccKey.json   # Firebase service account credentials
 ┗ 📜README.md            # This README file

Project Details

Firebase Setup

The system uses Firebase for storing and retrieving face entry and exit data. The setup involves:

• Firebase Admin SDK for authentication.
• Realtime Database for storing attendance records.
• Firebase Storage for storing entry and exit images.

Face Detection and Recognition

• Haar Cascade Classifier: Detects faces in the video feed.
• Face Encoding: Uses the face_recognition library to create unique encodings for detected faces.
• Face Matching: Compares the detected face encodings with known encodings to identify individuals.

Handling and Storing Data

1. Entry Logging: When a known face is detected, the system logs the entry time and uploads an entry image to Firebase Storage.
2. Exit Detection: Tracks how long a face remains visible and logs the exit time and image when the face disappears.
3. Data Recording: Stores the recorded information in the Realtime Database, including entry and exit images, timestamps, and attendance status (on time or late).

Contributing

If you'd like to contribute to this project, please fork the repository and use a feature branch. Pull requests are warmly welcome.