نظام حضور بالوجه

🧠 Facial Attendance System (Face Detection + Recognition + Fraud Detection)

The facial attendance system is a comprehensive, intelligent system for managing attendance and departure times. It utilizes artificial intelligence and computer vision technologies and aims to replace traditional methods such as handwritten signatures, cards, or fingerprints. The system uses a camera to capture and analyze faces in real time, then automatically decides whether to register or reject attendance based on several processing stages.

This system is characterized by its local operation using an SQLite database, making it lightweight and fast. It does not require an external server or a constant internet connection, making it suitable for small and medium-sized enterprises (SMEs) such as schools, companies, and institutes.

⚙️ System's Main Stages

The system relies on three main AI models that work sequentially and in an integrated manner:

1️⃣ Face Detection Model

This is the first stage of the system and is responsible for detecting the presence of any face within the camera's frame.

This model performs the following:

Analyzes the image or video in real time
Identifies the location of faces within the frame using bounding boxes
Isolates the face from the background
Enhances the image to prepare it for the next stage

Usually, advanced models are used at this stage, depending on the required level of accuracy.

The primary goal here is not to identify the person, but simply to answer the question: Is there a face in the image? And where is it located?

2️⃣ Face Recognition Model

After the face is detected, it is passed to the recognition model, which is the core of the system.

This model performs the following:

Extracts facial features and converts them into a digital fingerprint called Face Embedding
Represents the face as a mathematical vector containing hundreds of values
Compares this vector with data previously stored in an SQLite database

If the similarity between the new fingerprint and the stored fingerprints is very high (for example, more than 80% or according to the settings), the person is successfully identified.

This stage is responsible for answering the fundamental question: Who is this person?

3️⃣ Face Anti-Spoofing Model

This stage is one of the most important security phases in the system, as it prevents attempts to deceive the system.

The face anti-spoofing model verifies that the face is real and not fake, such as:

A printed image
A video on a phone or device screen
A 3D mask
A replay of an image from another device

The model relies on analyzing several factors, including:

Depth Information
Eye Movement and Blink Detection
Light Reflection on Skin
Texture Analysis
Facial Response to Natural Movement

Models such as CNN-based anti-spoofing or liveness detection models can be used for this purpose.

If a fake face is detected, the transaction is immediately rejected, and no attendance is recorded.

🗄️ SQLite Database

The system uses the SQLite database because it is lightweight, easy to integrate, and operates without complex configuration.

The database contains two main tables:

Users Table:
User ID
Full Name
Face Embedding
Additional data such as department or job title
Attendance Table:
Transaction Number
User ID
Date of Attendance
Entry and Exit Time
Verification Status (Real/Fake)

The data is stored locally in a single file, facilitating backup and transfer.

🔄 System Operation:
Real-time camera operation
Frame capture from video
Face detection model applied to locate the face
Face passed to forgery detection model for verification
If the face is genuine, data is sent to the recognition model
Face compared to the SQLite database
If a match is found, attendance is recorded with the time and date
Data is automatically saved to the database
🚀 System Features:
Operates automatically without human intervention
High accuracy in person recognition
Prevention of forgery or fraud attempts
Offline system
Sufficient real-time processing speed
Easy scalability and user addition
Suitable for schools, universities, companies, and institutions
🔐 Security and Reliability:

The system boasts a high level of security thanks to its forgery detection layer, preventing the use of images or videos to deceive the system. Furthermore, storing data in SQLite reduces the risk of external breaches, especially when the system is used locally.