VIHAN BHUYAN – INFORMATION RESEARCHER
#AI based attendance system
1) Source – International Journal of Advanced Research in Computer and Communication
Engine
The main goal of the proposed system is to:
Eliminate manual attendance marking
Prevent proxy attendance
The system works by using AI-based face detection and recognition techniques:
1. Image Capture
A camera (such as CCTV or a webcam) captures real-time images or video of
students/employees.
2. Face Detection
The system detects human faces from the captured image using computer vision
techniques (e.g., Haar Cascade classifiers).
3. Face Recognition
Detected faces are compared with pre-stored images in the database using machine
learning algorithms.
4. Attendance Marking
If a face matches with the database, attendance is automatically marked along with:
o Name / ID
o Date
o Time
5. Data Storage
Attendance records are stored in a centralized database for future access and
monitoring.
Technologies Used :
Artificial Intelligence (AI)
Machine Learning algorithms
Computer Vision
Digital Image Processing
Database Management Systems
AI-Based Smart Attendance Management System (IJISRT)
Purpose of the System
Traditional attendance methods are:
Time-consuming
Prone to human errors
Vulnerable to proxy attendance (someone marking for someone else)
The IJISRT system addresses these issues by using AI and computer vision to automate
detection, recognition, and attendance logging in real time, thereby improving accuracy,
speed, and reliability.
Haar Cascade Classifier (Face Detection)
The Haar Cascade / Viola-Jones framework is used to detect faces in each camera
frame by scanning for certain visual features like edges and textures.
It’s a machine-learning-based object detection method trained on large numbers of
face and non-face examples, so it can quickly identify where faces are in an image.
How an AI-Enabled Attendance System Works
An AI-enabled attendance system automatically records attendance by using Artificial
Intelligence (AI), machine learning, and computer vision technologies. Instead of manual
roll calls or biometric devices, it identifies individuals (usually through facial recognition)
and marks attendance digitally.
Overall Working Flow
Camera → Face Detection → Face Recognition → Attendance Marking → Data Storage
mage / Video Capture
A camera (webcam, CCTV, or mobile camera) is installed in the classroom or
workplace.
The camera captures:
o Live video stream, or
o Periodic images at fixed intervals
This captured data acts as the input for the AI system.
2️⃣ Face Detection (Who is present?)
The system first needs to find faces in the image.
AI uses computer vision algorithms such as:
o Haar Cascade Classifier
o HOG (Histogram of Oriented Gradients)
o Deep learning–based detectors (CNN, SSD)
How it works:
The algorithm scans the image frame.
It identifies regions that look like human faces.
Each detected face is isolated and cropped.
3️⃣ Face Pre-Processing
Before recognition, detected faces are prepared to improve accuracy:
Converted to grayscale
Resized to a standard dimension
Noise removal and normalization
Alignment (eyes and nose positioned correctly)
4️⃣ Face Recognition (Who is it?)
This is the core AI step.
🔹 Training Phase (Before system use)
Multiple images of each person are collected.
These images are labeled with:
o Name
o ID number
AI extracts unique facial features and stores them in a database.
🔹 Recognition Phase (During attendance)
The live face is compared with stored facial data.
Common algorithms:
o LBPH (Local Binary Patterns Histogram)
o FaceNet
o CNN-based deep learning models
If a match is found → Identity confirmed
If no match → Marked as unknown
5️⃣ Attendance Marking
Once a face is recognized:
The system automatically records:
o Name / ID
o Date
o Time
o Subject / Session (optional)
Duplicate entries are avoided using timestamps.
6️⃣ Data Storage and Management
Attendance data is stored in:
o Local database (MySQL, SQLite)
o Cloud database (optional)
Admins or teachers can:
o View attendance reports
o Export data (Excel / PDF)
o Monitor daily or monthly attendance
7️⃣ Monitoring & Reporting (Optional AI Features)
Advanced AI attendance systems may include:
Real-time dashboards
Attendance percentage calculation
Alerts for low attendance
Integration with Learning Management Systems (LMS)