🎯 Face Recognition Attendance System

A Python-based automated attendance management system that uses computer vision and machine learning to identify students through facial recognition and record their attendance with timestamps.

The system captures facial samples through a webcam, trains a K-Nearest Neighbors (KNN) classifier, recognizes registered students in real time, and stores daily attendance records in CSV files

📌 Project Overview

Traditional attendance systems require teachers to manually call names or maintain physical registers, which can be time-consuming and prone to human error.

This project automates the process using face detection and face recognition.

A student can register their face once, after which the system can identify them through a webcam and record their attendance along with the exact time.

Core Workflow

Student Registration
        ↓
Capture Face Samples
        ↓
Store Face Data
        ↓
Train KNN Classifier
        ↓
Real-Time Face Detection
        ↓
Face Recognition
        ↓
Student Identification
        ↓
Attendance Recorded
        ↓
Daily CSV Report

✨ Features

👤 Face Registration

The system allows a new student to register by entering their name and capturing facial samples through the webcam.

During registration:

Webcam captures the student's face
Haar Cascade detects the face
Face images are cropped and resized
100 facial samples are collected
Facial data is stored locally
Student names are associated with the captured samples

The registration process is implemented in add_faces.py.

🤖 Face Recognition

The system uses:

OpenCV for webcam input and face detection
Haar Cascade Classifier for detecting faces
KNN for identifying registered students

During recognition, the detected face is resized to the same dimensions used during training and passed to the KNN classifier to predict the student's identity.

🕐 Automated Attendance

Once a student is recognized, the system generates an attendance record containing:

NAME
TIME

Attendance is stored in a daily CSV file using the format:

Attendance/Attendance_DD-MM-YYYY.csv

This creates a separate attendance file for each day.

🔊 Voice Confirmation

The system uses Windows Speech API through win32com.client to provide voice feedback when attendance is taken.

For example:

"Attendance Taken.."

This gives the user an audio confirmation after pressing the attendance key.

📊 Attendance Dashboard

A Streamlit-based application is included for displaying the attendance data.

The application automatically loads the CSV corresponding to the current date and displays the records using a Streamlit dataframe.

🛠️ Tech Stack

Technology	Purpose
Python	Core programming language
OpenCV	Webcam processing and computer vision
Haar Cascade	Face detection
Scikit-learn	KNN machine-learning classifier
NumPy	Numerical and matrix operations
Pandas	Attendance data handling
Pickle	Local face/label data storage
CSV	Attendance record storage
Streamlit	Attendance dashboard
PyWin32	Voice feedback through Windows SAPI
HTML/CSS/JavaScript	Basic attendance portal interface

The repository includes Python scripts for registration, recognition, testing, and the Streamlit viewer, along with a basic web portal.

📂 Project Structure

Attendence-System/
│
├── Attendance/
│   └── Attendance_DD-MM-YYYY.csv
│
├── data/
│   ├── haarcascade_frontalface_default.xml
│   ├── names.pkl
│   └── faces_data.pkl
│
├── Attendance/
│
├── Login.html
├── main.html
├── script.js
├── stylee.css
│
├── add_faces.py
├── app.py
├── test.py
│
├── xyz.jpg
├── xyzz.png
│
└── README.md

The current repository contains the Attendance and data directories along with the Python scripts and web-interface files shown above.

⚙️ How It Works

Step 1 — Register a Student

Run:

python add_faces.py

Enter the student's name when prompted:

Enter Your Name:

The webcam opens and captures up to 100 face samples for the student.

The captured data is stored in:

data/faces_data.pkl

and the corresponding names are stored in:

data/names.pkl

Step 2 — Start Face Recognition

Run:

python test.py

The program:

Opens the webcam
Detects faces
Loads registered face data
Trains a KNN classifier
Predicts the detected student's name
Displays the recognized name on the video feed
Records attendance when the attendance key is pressed

The current implementation uses:

KNeighborsClassifier(n_neighbors=5)

for classification.

Step 3 — Mark Attendance

When the student's face is recognized, press:

The system provides voice confirmation and writes the student's name and timestamp into the day's attendance CSV file.

Press:

to exit the recognition window.

Step 4 — View Attendance

Run the Streamlit application:

streamlit run app.py

The dashboard loads the current day's attendance CSV and displays the records in a table.

🧠 Machine Learning Approach

The project uses a K-Nearest Neighbors (KNN) classifier.

Training

Registered face images are:

Face Image
    ↓
Face Detection
    ↓
Crop Face
    ↓
Resize → 50 × 50
    ↓
Flatten Pixel Values
    ↓
KNN Training Data

The training data and labels are loaded from:

faces_data.pkl
names.pkl

The KNN model is configured with:

KNeighborsClassifier(n_neighbors=5)

and trained using the stored face matrix and corresponding labels.

📸 Face Detection

The project uses OpenCV's Haar Cascade classifier:

data/haarcascade_frontalface_default.xml

The webcam frame is converted to grayscale before face detection:

gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)

The detected face is then cropped and resized before being passed to the recognition pipeline.

📊 Attendance Data Format

Attendance is stored as CSV files:

Attendance/
    ├── Attendance_15-08-2026.csv
    ├── Attendance_16-08-2026.csv
    └── Attendance_17-08-2026.csv

Each file contains:

NAME,TIME
Lakshya,10:30-15
Rahul,10:32-08
Aman,10:34-21

This makes the records easy to open and analyze using Excel, Pandas, or other data-analysis tools.

🖥️ Basic Web Portal

The repository also contains a basic attendance portal built with HTML, CSS, and JavaScript.

The portal provides two primary actions:

┌──────────────────────────────┐
│      Attendance Portal       │
│                              │
│   [ Mark Attendance ]        │
│                              │
│   [ New Registration ]       │
└──────────────────────────────┘

The interface is implemented through main.html, script.js, and stylee.css.

🚀 Installation

Prerequisites

Make sure you have:

Python 3.x
Webcam
Windows OS recommended for voice feedback
Git

Clone the Repository

git clone https://github.com/Lakshya172/Attendence-System.git

Navigate to the project:

cd Attendence-System

Install Dependencies

pip install opencv-python
pip install numpy
pip install pandas
pip install scikit-learn
pip install streamlit
pip install streamlit-autorefresh
pip install pywin32

▶️ Running the Project

Register a Student

python add_faces.py

Start Recognition

python test.py

Start Attendance Dashboard

streamlit run app.py

⚠️ Important Notes

Camera Required

The system uses:

cv2.VideoCapture(0)

so a working webcam is required.

Windows Voice Support

The current recognition script uses:

from win32com.client import Dispatch

for Windows SAPI voice output, so this feature is intended for Windows environments.

Local Data Storage

Face data is stored locally using Pickle files:

data/names.pkl
data/faces_data.pkl

Attendance records are stored as CSV files rather than in a cloud database.

🔮 Future Improvements

The project can be extended into a production-ready attendance platform by adding:

🎓 Learning Outcomes

Through this project, I gained practical experience with:

Computer Vision
Face Detection
Face Recognition
Machine Learning Classification
OpenCV
KNN Algorithm
Webcam Processing
Python File Handling
CSV Data Management
Pickle Serialization
Streamlit
Basic Web Development
Real-time Application Development

👨‍💻 Author

Lakshya Agarwal

B.Tech Computer Science & Engineering Lovely Professional University

GitHub: https://github.com/Lakshya172

⭐ Project

If you find this project useful, consider giving the repository a ⭐.

Repository: https://github.com/Lakshya172/Attendence-System

📄 License

This project was created for educational and learning purposes.

Name		Name	Last commit message	Last commit date
Attendance		Attendance
data		data
Login.html		Login.html
README.md		README.md
add_faces.py		add_faces.py
app.py		app.py
main.html		main.html
script.js		script.js
stylee.css		stylee.css
test.py		test.py
xyz.jpg		xyz.jpg
xyzz.png		xyzz.png

Latest commit

History

Folders and files

Repository files navigation