L. D.

COLLEGE OF ENGINEERING

A Project Report on

ASFR: Attendance System by Face Recognition

Design Engineering -2B (Subject Code: 3160001)

B.E. Semester 5
Information Technology

NAME ENROLLMENT NO.

1. SEJU AMIT 220280116119

2. SANCHANIYA SUJAL 220280116114

3. VASAVA ABHAY 220280116149

4. KHAMAR PRIYANSHU 220280116045

Internal Faculty Guide:

Pro. Shital Solanki
 DEPARTMENT OF
INFORMATION & TECHNOLOGY

L. D. COLLEGE OF ENGINEERING
AHMEDABAD

CERTIFICATE
This is to certify that the basic process and framework has been satisfactorily
carried out by,

NAME ENROLLMENT NO.

1. SEJU AMIT 220280116119
2. SANCHANIYA SUJAL 220280116114
3. VASAVA ABHAY 220280116149
4. KHAMAR PRIYANSHU 220280116045

Under my guidance in the fulfillment of the course DESIGN

ENGINEERING during the academic year 2024-25.

Date of Submission:

Pro. Shital Solanki Dr. Hiteishi Diwanji

(Faculty Guide) (H. O. D.)

 Acknowledgement

I wish to express my heartfelt appreciation to all those who have contributed to this project, both
explicitly and implicitly, without the cooperation of whom, it would not have been possible to complete
this project. This second year project has been carried out at the Department of Chemical, Gujarat
Technological University. I would also like to thank my internal guide Prof Shital Solanki, The faculty
members of LDCE who had helped me directly or indirectly in my project work.

We are also deeply thankful to Dr. Hiteishi Diwanji, our Head of Department, for her
constant encouragement and for creating an environment that fosters learning and innovation.
Her support and leadership were instrumental in helping us navigate challenges and stay
focused on our goals. Additionally, we would like to thank L.D. College of Engineering for
the opportunities and support provided during this project. I thank my parents for standing
behind me all the time and I thank my friends for their advice, guidance and help. I thank
God for all his blessings.

3
 Abstract
All institutions in our country are concerned about regularity of attendance. Mainly there are two conventional
methods for attendance taking and they are by calling names or by taking sign on paper. They both were more
time consuming and inefficient method. Hence, there is a requirement of computer-based attendance
management system which will assist for maintaining attendance of presence. This project defines the basic
problem of attendance management in various organization. In this project the various computerized system
which is being developed by using different techniques. Based on this a new approach for attendance recording
and management is proposed to be used for various organization.

INTRODUCTION

The technology aims in imparting a tremendous knowledge oriented technical innovation these days. Deep
Learning is one among the interesting domain that enables the machine to train itself by supplying some
datasets as input and supplies a suitable output during testing by applying different learning algorithms.
Nowadays Attendance is considered as a key factor for Industries as well as an educational organization. With
the advancement of the deep learning technology the machine automatically detects the attendance
performance and keeps a record of those collected data. In general, the attendance system can be kept in two
different forms namely,

Manual Attendance System (MAS)

Automated Attendance System (AAS).

Manual' Attendance Management system is a process where an Industries as well as an educational

organization concerned with the subject need to call the name and mark the attendance manually. Manual
attendance may be considered as a time-consuming process or sometimes it happens for the teacher to miss
someone or students may answer multiple times on the absence of their friends. So, the problem arises when
we think about the traditional process of taking attendance in the classroom. To solve all these issues, we go
with Automatic Attendance System (AAS).

Automated Attendance System (AAS) is a process to automatically estimate the presence or the absence of the
student in the classroom by using face recognition technology. It is also possible to recognize whether the
student is sleeping or awake during the lecture and it can also be implemented in the exam sessions to ensure
the presence of the student. The presence of the students can be figured out by capturing their faces on to a
high-definition monitor video streaming service, so it becomes reliable for the machine to understand the
presence of all the students in the classroom. The two common Human Face Recognition techniques are,

Feature-based approach

Brightness-based approach.

4
 INDEX

Topics Sr no.

Introduction 1
Design thinking phase 2
AEIOU Canvas 3
Mind Mapping 4
Empathy Phase 5
Ideation Phase 6
Product development 7
References 8
Appendix 9
Advantage and Disadvantage 10

5
Introduction
Efficient attendance tracking is a fundamental requirement for educational institutions,
corporate offices, and various organizations. Traditional attendance systems that rely on
manual sign-ins, paper sheets, or swipe cards have proven to be time-consuming, vulnerable
to manipulation, and susceptible to human error. These challenges often lead to inaccurate
records, impacting productivity, security, and accountability within the organization.

The Face Recognition Attendance System leverages cutting-edge computer vision and
artificial intelligence (AI) technologies to provide a modern, automated solution for
attendance management. By using facial recognition, the system identifies individuals based
on unique facial features, enabling hands-free, fast, and secure attendance tracking. With
minimal human intervention, this approach offers significant advantages, such as reducing
administrative workload, preventing proxy attendance, and increasing overall accuracy and
reliability.

Objectives of the Project

This project aims to:

 Develop an automated attendance system that recognizes faces in real-time, enhancing both
security and convenience.
 Integrate machine learning and computer vision techniques to improve recognition accuracy
under various conditions (lighting, pose variations, etc.).
 Create a user-friendly interface that simplifies attendance management, data retrieval, and
reporting.
 Ensure privacy and security by securely storing and processing biometric data.

Expected Impact and Benefits

Implementing a face recognition-based attendance system can significantly streamline

attendance processes, saving time and reducing costs associated with manual or card-based
systems. The system also contributes to a secure environment by ensuring only authorized
individuals are marked present, which is particularly useful in environments requiring strict
access control.

This technology is adaptable and scalable, allowing it to be deployed in diverse

environments, from classrooms and corporate offices to restricted access areas. As AI and
facial recognition technologies advance, the system can further evolve to improve recognition
accuracy, adapt to a wider range of user conditions, and enhance overall efficiency.

In summary, the Face Recognition Attendance System represents a transformative shift in

attendance management, harnessing advanced technology to address longstanding issues in a
practical and forward-looking manner. This report discusses the design, development, and
evaluation of the system, providing insights into its potential to redefine how attendance is
managed across various domains.

6
CANVAS

2. Design and Thinking Phase

2.0 AEIOU Canvas
Evaluation of idea:

AEIOU charts describe overall ideas about all details about previous concrete.

This comprises of five elements

1. Activities: the general observation or activities done in the domain.

2. Environment; the activities done in a specific surrounding is called environment.

3. Interaction: all type of interaction performed to get the job done.

4. Object: all the props, tools or objects used to get the job done.

5. Users: all the individuals that took part in the process are users.

2.0.1 Activities

These are goal-directed sets of action towards things people want to accomplish. What
are the modes people work in, and the specific activities and processes they go through?

In Summary of activities We describe general Impression, elements and summary of activities in

general impression we describe what is happening on site with people. Here are some activities

• Attendee Registration

• Photo Capture

• Image Extraction

• Photo comparison in database

• Attendance marking in excel

• If absent SMS or Email

2.0.2 Environment

It includes the entire arena where activities take place. In Summary of ENVIRONMENT.

We describe general impressions, elements, and features.

In general, we describe the surroundings.

7
 • Indoor areas with sufficient light

• Clear face without any covers

• Works under all temperature condition Suitable for all working condition

2.0.3 Interaction

These are between a person and someone or something else; they are the building block of
activities,In Summary of INTERACTIONS, we describe interactions that how people should interact
with each other

Here we describe general impression and elements

• Teacher-Student Boss-Employee

• Principal-Teacher

• Higher to lower authorities

2.0.4 Object

These are between a person and someone or something else. They are the building blocks of
the environment of activities.

In Summary of INTERACTIONS

We describe interactions that how people should interact with each other

Here we describe general impression and elements

• Camera

• Storage device

• Computer

• Power supply

• Registration application

2.0.5 Users

Users are the people whose behavior's, preferences, and needs are being observed,In Summary
of USERS, we describe in general impression and elements.In general impression we describe this as
the individual who participates in the process.

• School

• Colleges

• Factories

8
9
2.1 Mind Mapping
Mind mapping gives us the detailed information of how the project works with all its
participants, dependencies and the objects used.

10
2.2 Empathy Phase
2.2.1 Users
In an empathy canvas it shows you the details or the basic summary of multiple behavioral
relation between different objects in this case users have different empathetic relations like.

• Educational institutes

• Banks

• Airport authorities

• Security services

2.2.2 Stakeholders
This section defines the different empathetic behaviors of stakeholders in that particular
domain.This could be an individual or group of individuals or even some corporation. These are
below mentioned stakeholders.

• Engineers

• Dealers

• Suppliers

• Investor

2.2.3 Activities
This section of the empathy canvas determines all the activities performed by all participants.

Activities are as listed below:

• Person registration

• Face capturing by camera

• Photo extraction and processing

• Face compression with database

• Person found or attendance marked

• Absent or present data stored in excel Alert through SMS or Email

2.2.4 Story Boarding

HAPPY: We used this project in our college classroom. As it uses facial recognition for
attendance so there was no requirement of taking attendance manually thus making it smart
and time saving
11
HAPPY: There is no need of manually counting the attendance of each student at the end of
semester as the prototype does it automatically by maintain the computerized records of
each student separately in its database.

SAD: As our prototype pre stores the personal information like name, age, enrollment no.. etc.
of student and employee so there is big chance of data breach through cyber- attack and also
issue of privacy needs to be addressed.

SAD: All the attendance data is stored in our database of storage device. If any case of male
function of a storage device. All the important attendance data can be lost.

12
2.3 Ideation Phase
2.3.1 People: This is software related to technical as well as education fields so it affects the
people attached to it.Some of the people mentioned below are affected indirectly.

• Teacher

• Professor

• Manger

• Principal

• Organizer

• Pollice

• Security personals

2.3.2 Activities: Here people are engaged in some other activities some of us are mentioned
below:

• Daily attendance

• Reduce manual work significantly

• Monthly and yearly attendance report

• Data analysis of attendance report SMS or Email if absent

2.3.3 Situation/Context/Location: Then thought for context/location/situation and

finally for possible solutions. We conclude that people's emotions when using an interactive system
do not only influence the current interaction itself.

• Taking attendance

• Finding persona in database

• For time efficient and smart administration

• Educational institutes

• Airports and Railway stations

• Banking sector

• Security and defense services

2.3.4 Props/Tools/Objects:
Camera Database handler
13
• Raspberry pi

• Python

• Display

• Storage device

• Safety covers

14
2.4 Product Development Phase
1. Purpose
• To develop a time efficient attendance system

• Excluding the manual work

• To develop a secure attendance system

2. People
• Students

• Teacher

• Banking staff

• Professor

• Airport authorities

3. Product experience

• 24*7 working

• Time efficient Zero error system

4. Product functions

• Face recognition

• Comparing face with database

• Alert through SMS or Email

5. Product features

• Quick response

• High sensitivity

• High security

• Cost effective solution

6. Components

• Arduino

• OLED display

15
• Micro SD card Raspberry Pi 4

• Cameras

• SD card modules

• Cabels

16
2.5 Learning Needs Matrix
2.5.1 ) Purpose/Product concept :

Purpose

1. Automation:
o To automate the attendance-taking process, reducing the time and effort
required to manually mark attendance.
2. Accuracy:
o To enhance the accuracy of attendance records by minimizing human error
and ensuring that the right individuals are marked present.
3. Convenience:
o To provide a user-friendly and efficient method for checking in without the
need for physical ID cards, roll calls, or manual entry.

2.5.2 ) Tools/Methods/Theories Application Process Involved

During II :
Requirement Analysis

 Identify Use Case: Determine the specific context (e.g., school, corporate) and the needs of
users (students, employees).
 Gather Requirements: Understand features needed, such as real-time tracking, reporting,
security requirements, and integration with existing systems.

During III :
System Design

 Architecture Design: Plan the system architecture, including client-server interactions,

database design, and user interface.
 User Interface (UI) Design: Create wireframes and prototypes of the application for better
user experience.

During IV :
Technology Selection

 Select Tools and Frameworks: Choose programming languages (e.g., Python for backend,
JavaScript for frontend), frameworks (e.g., Flask, React), and databases (e.g., MySQL,
MongoDB).
 Choose Face Recognition Libraries: Select suitable libraries for face recognition, such as:
o OpenCV: For image processing.
o Dlib: For face detection and facial landmark identification.
o Face_recognition: A simple library built on top of Dlib for face recognition tasks.
17
2.5.3 ) Applicable stanars and design specification/Principles and
Experiments

During II :

1. Applicable Standards

Data Protection Standards

 GDPR (General Data Protection Regulation): For applications operating in the European
Union or processing EU residents' data, compliance with GDPR is essential. This includes
obtaining user consent, ensuring data protection rights, and implementing privacy by design.
 CCPA (California Consumer Privacy Act): For applications targeting users in California,
compliance with CCPA involves transparency about data collection practices and user rights
regarding their personal data.
 ISO/IEC 27001: A standard for information security management systems (ISMS), ensuring
that user data is securely handled and protected.

During III :

Design Specifications/Principles

User-Centric Design

 Intuitive Interface: The user interface should be simple and intuitive, enabling users to easily
navigate through the system and perform tasks like checking attendance or registering.
 Feedback Mechanisms: Provide immediate feedback to users upon successful face
recognition or errors, enhancing user experience.

Security Principles

 Data Encryption: Use strong encryption algorithms to secure sensitive user data, both in
transit (using HTTPS) and at rest.
 Access Control: Implement role-based access controls to ensure that only authorized users
can access specific features or data within the system.

During IV :

Experimental Considerations

Performance Testing

 Accuracy Testing: Conduct experiments to evaluate the accuracy of the face recognition
algorithm under various conditions (lighting, angle, occlusion). Metrics like precision, recall,
and F1-score can be used to quantify performance.
18
  Load Testing: Simulate multiple users logging in simultaneously to assess how the system
performs under high loads and identify any bottlenecks.

2.5.4 ) Software/ Simulation / Skill / Mathematical Requirement

During II :
Programming Languages

 Python: Primarily for backend development, implementing machine learning algorithms, and
face recognition tasks.
 JavaScript: For frontend development, creating interactive web applications.

During III :
Libraries and Frameworks

 Face Recognition Libraries:

o OpenCV: For image processing, face detection, and recognition.
o Dlib: For implementing facial landmark detection and recognition.
o Face_recognition: A simple library built on top of Dlib for face recognition tasks.
 Machine Learning Frameworks:
o TensorFlow or PyTorch: For building and training deep learning models, especially if
using CNNs for face recognition.

During IV :
Development Environments

 Integrated Development Environments (IDEs):

o PyCharm or VS Code: For Python development.
o WebStorm: For JavaScript and frontend development.

B. Simulation and Testing Tools

 Postman: For testing API endpoints and ensuring that the backend works as expected.
 Jupyter Notebooks: For experimenting with machine learning models and data analysis.
 Anaconda: A distribution for managing Python packages and environments, useful for
machine learning and data science projects.

2.5.5 ) Components materials’ strength criteria ( exploration-

varieties/ testing requirement )

During II :

Hardware Components and Strength Criteria

19
A. Cameras

The camera is one of the most critical hardware components of the attendance system,
responsible for capturing high-quality facial images.

Varieties:

 Webcams: Low-cost, widely available, used for individual workstations or smaller setups.
 IP Cameras: Network-based, used for real-time monitoring in larger setups like classrooms
or offices.
 CCTV Cameras with Facial Recognition: Used in large, secure environments where facial
recognition is integrated with surveillance.
 Infrared Cameras: For low-light environments, providing higher accuracy by detecting heat
signatures and avoiding lighting dependency.

During III :
Processing Unit (Edge Computing Devices)

Edge computing devices are used to process facial data locally, reducing latency and reliance
on cloud infrastructure.

Varieties:

 Single-board Computers (e.g., Raspberry Pi): Used for smaller installations or where cost
efficiency is key.
 Dedicated GPU Workstations: For processing large volumes of facial data in real-time,
particularly useful in environments where accuracy and speed are critical.
 Edge AI Devices: These are purpose-built for real-time AI processing, such as Google Coral or
NVIDIA Jetson.

During IV :
Network Infrastructure

The network infrastructure ensures the transfer of data between cameras, processing units,
and the central server for attendance logging.

Varieties:

 Wired Networks (Ethernet): Provides stable, high-speed connections, reducing latency.

 Wireless Networks (Wi-Fi): Used where wired connections are impractical, but may
introduce higher latency or network congestion.

Strength Criteria:

 Bandwidth: The network should support high-speed data transfer, particularly for systems
with high-resolution video feeds or multiple cameras.
 Latency: Low latency is crucial for real-time attendance tracking.
20
 Reliability: The network must remain reliable even under peak loads.

21
Advantages & Disadvantages :
Automated time tracking system

Offices or workplaces or even just public places where the entry and exit times of employees or a
person are strictly noted down will have a ready-made automated system to record the entry and
exit time of each person for a given time. It will not even need the person to stop and click a photo,
the software's are advanced enough to record the data from a continuous reel also. This means the
flow will not get hampered, or you will not have to stop and smile or something like that. Just enter
or exit the place effortlessly like you do every day and boom! Your attendance will be recorded
without any fuss!

Cost-effective

Since the entire process will be done by a computer, it means the total attendance registration and
calculation will be automated and done by the system itself, therefore, saving us the money which
would have been otherwise spent on the labor cost to do that.

Increased security

Face recognition-based attendance system will not just calculate attendance but also note down the
entry and exits of visitors in the place. At times when there is a situation where the identity and time
of entry and exit of a specific person need to be noted, this system would become handy as it will
easily show you when he/she came in and what are the places he/she went to a very precise level.
All of this means, you will have a much higher security level in your workplace

Time saving

The entire world is suffering from COVID19 and it is high time we must give heed to social distancing.
Having a safe distance with others has become a necessity nowadays. Times like this can be
problematic if you have manual attendance system, having a Face recognition-based attendance
system will not only allow you to register the attendance of a person but also keep you at a safe
distance from them as you can work remotely and still see who all are coming and going. This calls
for the point that, this entire system is a much safer, time-saving, and faster method to record
attendance.

Easy to manage

Since the artificial intelligence-based attendance system is fully automated, managing the records
and keeping a track of day-to-day activities will become much easier than the manual system.
Everything will be done by the system. Many software is programmed in such a way that it shows
the exact time of how many hours or minutes a person worked on his/her desk in the day. All this is
can be done on an exceptionally large scale. Just imagine, recording the activities of a large crowd of
200 people simultaneously without any fuss and recording it at the same time in an organized
manner. Such is the power of Al in face recognition.

Disadvantages

22
If in the wrong hands, it will be a disasterJust like we have already read that, the touchless biometric
attendance system will record everything and is a handy method to analyze and register a person's
activities, if in wrong hands can cause a serious level of the privacy breach. Governments in some of
the most treacherous countries in the world use this technology to analyze the activities of their
citizens. Which means you will not be able to go anywhere without the government knowing. Cases,
where a particular person or a group uses contactless biometric attendance, will mean, they can
easily watch the whereabouts of a person meaning you will be always tracked. This means that Al in
Face recognition system can be easily worked against us.

Data privacy breach

Data in these systems means, billions of pictures and God knows. 1000s of hours of video footage
saved into a hard drive. This calls for the point when this hard drive gets hacked or smuggled to get
data of a particular person. The data can be easily tampered and used to in against a particular
organization or person. Which can be very much dangerous. Therefore, companies that use Artificial
intelligence-based attendance system have an extremely strict security to safeguard the information
of their employees.

Low reliability

Sometimes there have been instances where the identity of a person is not able to get verified.
There have been also cases where the identity of a person is verified with another person's identity.
This means a person who is "X" is recognized as "Y" instead of "X." Times like this can be misused as
criminals can loot or commit a crime in the name of other people. This means. that even though we
have achieved a lot of advancement in the technology sector, there will always be a small gap
between us and the ideal system.

Lack of regulations in the AI in Face recognition systems

We can already guess how a particular Artificial intelligence-based attendance system can be
misused. In addition to all these, the govt across the entire world do not have a particular policy or
regulations to regulate the usage of touchless biometric attendance system. Some cities and places
in the world have already banned this system considering the amount of threat it can cause to the
people. While there are some places in the world where usage of face recognition-based attendance
system is already under progress but without any regulation. Which means lakhs of people's data is
already under threat.

Keeping all these things in mind, the advantages of Artificial intelligence-based attendance system
are many, but so are the disadvantages. But we can see that this system will be much better if all the
processes are regulated and taken care of. The first point will be: -

Proper rules and regulations while setting up and using touchless biometric attendance system.
Companies or organizations should be warned on what level the information of people can be used
or shared. Keeping the sharing freedom to the least extent possible, the government caneasily
consult experts and set up policies particularly based on this Artificial intelligence-based attendance
systems.

23
Secondly, companies or organizations should pay much attention to the security of the information
rather than other things in setting up this system. The major setup that this whole contactless
biometric attendance will need is a small camera at the entry and exit points of the rooms and a
computer and a storage device to store the data. All this will not take more than 2×2" space. So,
keeping these things aside, people should be much more advanced and stricter when it comes to the
security of this system.

Thirdly, face recognition-based attendance system should be worked increasingly every day to take
it closer to the ideal system that we need. The algorithms should be worked upon and data set that
is feed initially to the system should be much larger to remove glitches. Keeping in mind, the glitches
that this system can cause can be much disastrous in situations where a person gets paid on a "no
work no pay" policy job. He/she may get marked absent on a particular day even if he had come, just
because the contactless biometric attendance system was faulty and not correct enough to register
his face. At police stations or courts where the identity of a person is identified using this
technology, it may cause some serious trouble if a particular person is identified as someone else in
the records. This will not only create a problem but also allow criminals to make use of this flaw
every time they commit a crime.

24