Dissertation on

“Attendance Monitoring System Through RFID and Face Detection”

Submitted in partial fulfilment of the requirements for the award of degree of

Bachelor of Technology
in
Computer Science & Engineering
UE21CS390A – Capstone Project Phase - 1
Submitted by:

Kapil N PESUG21CS320
Pratham CS PES1UG21CS441
Vaibhav M PES2UG21CS585
Suhrid Sen PES2UG21CS635

Under the guidance of

Prof. Priya Badrinath

Asst.Professor

January - May 2024

DEPARTMENT OF COMPUTER SCIENCE AND ENGINEERING
FACULTY OF ENGINEERING
PES UNIVERSITY
(Established under Karnataka Act No. 16 of 2013)
100ft Ring Road Bengaluru – 560 085 Karnataka India
 PES UNIVERSITY
(Established under Karnataka Act No. 16 of 2013)

100 Feet Ring Road Bengaluru – 560 085 Karnataka India

FACULTY OF ENGINEERING

CERTIFICATE
This is to certify that the dissertation entitled

Attendance Monitoring System Through RFID and Face Detection

is a bonafide work carried out by

Kapil N PES2UG21CS320
Pratham CS PES1UG21CS441
Vaibhav M PES2UG21CS585
Suhrid Sen PES2UG21CS635

in partial fulfilment for the completion of sixth semester Capstone Project Phase - 1 (UE21CS390A) in the Program of
Study - Bachelor of Technology in Computer Science and Engineering under rules and regulations of PES University
Bengaluru during the period Jan. 2024 – May. 2024. It is certified that all corrections / suggestions indicated for internal
assessment have been incorporated in the report. The dissertation has been approved as it satisfies the 6 th semester academic
requirements in respect of project work.

External Viva
Signature Signature Signature
<Name of the Guide> Dr. Mamatha H R Dr. B K Keshavan
Designation Chairperson Dean of Faculty

Name of the Examiners Signature with Date

1. __________________________ __________________________

2. __________________________ __________________________
 DECLARATION

We hereby declare that the Capstone Project Phase - 1 entitled “Attendance Monitoring
System Through RFID and Face Detection” has been carried out by us under the
guidance of Prof. Priya Badrinath Asst.Professor and submitted in partial fulfilment of the
course requirements for the award of degree of Bachelor of Technology in Computer Science
and Engineering of PES University Bengaluru during the academic semester January –
May 2024. The matter embodied in this report has not been submitted to any other University
or Institution for the award of any degree.

Kapil N PES2UG21CS320 ________________________

Pratham CS PES2UG21CS441 ________________________

Vaibhav M PES2UG21CS55 ________________________

Suhrid Sen PES2UG21CS635 ________________________

ACKNOWLEDGEMENT

We would like to express our gratitude to Prof. Priya Badrinath Department of Computer
Science and Engineering PES University for her continuous guidance assistance and
encouragement throughout the development of this UE19CS390B – Capstone Project Phase – 1.

We are grateful to the project coordinator Dr Priyanka H all the panel members & the supporting
staff for organizing managing and helping the entire process.

We take this opportunity to thank Dr. Mamatha H R Chairperson Department of Computer Science
and Engineering PES University for all the knowledge and support we have received from her.

We are grateful to Dr. M. R. Doreswamy Chancellor PES University Prof. Jawahar Doreswamy
Pro Chancellor – PES University Dr. Suryaprasad J Vice–Chancellor Dr. B.K. Keshavan Dean
of Faculty PES University for providing us various opportunities and enlightenment during every
step of the way.

Finally this project could not have been completed without the continual support and
encouragement we have received from our family and friends.
 ABSTRACT

The objective of the capstone to develop a system that can revolutionize the
existing attendance management system by combining real-time face recognition
and RFID technology. The use of paper sheets or swipe cards can be erroneous,
time-consuming and not real-time. The developed system will solve these
limitations by providing an accurate, prompt, and easy-to-use attendance
management solution for workspace, event-working places and academic
premises. The system will take the massive strides in terms of technology to
develop a facial recognition and use RFID and reporting in real-time to the
attendance data.
 TABLE OF CONTENT

Chapter Title Page

No. No.
1. INTRODUCTION 1
2. PROBLEM STATEMENT 2
3. LITERATURE REVIEW 3
3.1 Face Recognition And Detection 3
3.1.1. Real-Time Face Recognition System (RTFRS) 3
3.1.2. Real-Time Face Detection and Segmentation 4
3.1.3. Frontal Face Generation Based Multi-angle Face Identification System 5
3.1.4. Design of Face Detection and Recognition System for Smart Home Security
6
Application
3.1.5. Multi-Faces Recognition Process Using Haar Cascades and Eigenface
6
Methods
3.1.6. Multiple Face Detection Attendance System 7
3.2. RFID Technology 8
3.2.1. The RFID Technology and its Applications: A Review 8
3.2.2. RFID Technology and Its Applications in Internet of Things (IOT) 10
3.2.3. RFID Based Attendance System 11
4. DATA 12
5. SYSTEM REQUIREMENTS SPECIFICATION 13
5.1 System Overview
5.2 User Roles
5.2.1 System Administrator
5.2.2 Registered User
5.3 Functional Requirements
5.3.1 User Management
5.3.2 RFID Tag Management
5.3.3 Real-Time Face Recognition
5.3.4 Attendance Marking
5.3.5 Reporting
5.4 Non-Functional Requirements
5.4.1 Performance
5.4.2 Security
5.4.3 Usability
5.4.4 Reliability
 5.4.5 Maintainability
5.4.6 Scalability
5.5 Additional Considerations
6. SYSTEM DESIGN (detailed)

6.1 Server
6.2 Database
6.3 RFID Scanner
6.4 Raspberry Pi Camera
7. IMPLEMENTATION AND PSEUDOCODE (if applicable)

8. CONCLUSION OF CAPSTONE PROJECT PHASE-1

8.1 Key Accomplishments

8.1.1 Comprehensive Literature Review
8.1.2 System Architecture Definition
8.1.3 Hardware Component Selection
8.2 Conclusion
9. PLAN OF WORK FOR CAPSTONE PROJECT PHASE-2

9.1 Refining the System Design

9.2 Software Development
9.3 System Integration and Testing
9.4Deployment Planning
REFERENCE/ BIBLIOGRAPHY

LIST OF FIGURES
Figure No. Title Page No.

Figure 1 System Design for Attendance marking using RFID and RTFR 21
Attendance Monitoring System Through RFID and Face Detection

Chapter 1
INTRODUCTION

The current state of attendance management is time taking and requires more
human efforts. Manual attendance sheets are inconvenient prone to inaccuracies
and susceptible to manipulation. In this day prevalence, a much more reliable
and secure solution is needed. This project intends to present this solution: an
Automated Attendance Management System, which ingeniously unites Real-
Time Face Recognition and Radio Frequency Identification . Picture a system
where: The Outcome The Persurrection of user identification and recording
unauthorized attendance filing are over. This device will verify the user’s
presence through RFID and flash an image using RTFR . The Outcome Proxy
attendance will also become a thing of the past. The device will check the user’s
physical presence via RFID and his unique facial signature using RTFR . This
barrier to proxy attendance will be almost impassable. Outcome Manually
processes will be eliminated. It will free up time for administrators and users,
and It can be done In a swipe of an RFID tag and a snap of the face. Outcome
No waiting for reports and a compilation of data. This will provide the
administrator with accurate attendance data for real-time performance analysis,
Decision making, and addressing the situation if need be. This project focuses
on creating and implementing a new system. By combining the power of RTFR
and RFID, it aim for a sweeping reform of attendance filing achievement of
results. This introduction offers a clearer understanding of the project’s thesis
and provides a glimpse into the difficulties it is designed to overcome. The
introduction also elaborates on the benefits and sets the framework for a novel
solution.
.

Dept of CSE Jan – May P a g e 1 | 29

Attendance Monitoring System Through RFID and Face Detection

Chapter 2
PROBLEM STATEMENT

The current Schools and colleges attendance management system has numerous
weaknesses that affect the efficiency, accuracy, and security of the system.
Efficiency. The current system, which is frequently paper-based or manual, is
labor-intensive for managers and users. Staff or teachers are forced to take time
off from their primary responsibilities as they distribute, collect, and enter the
data . Users are also faced with delays during peak intervals, which result in
crowding and irritation. Inaccuracy. The existing data entry through a paper-
based system is error-prone, leading to probable inconsistencies as well as
unreliable hops. Data-entry incidents can result in losing the attendance of an
employee or an incorrect tally, which requires time-consuming reconciliation.
Security vulnerabilities. In addition to external fraud generating false
attendance, traditional attendance systems are unable to safeguard against buddy
punching, which is when anyone else marks another’s attendance or students
use proxy attendance. The current system’s weaknesses have the following
repercussions: Wasting time. The current system’s inefficiencies result in the
loss of time for both administrators and users Frustration. Users and
administrators become irritated because of delays in applying or inaccuracy
Proving unreliable data. An unreliable hops record impedes informed decision
and can result in payroll issues or a failure to discipline Security vulnerabilities.
Bundy punching and proxy mating jeopardize the accuracy of hops entries
which raises security concerns. As a result, it is critical to create a more
efficient, accurate, and secure hops management system in schools and colleges .

Dept of CSE Jan – May P a g e 2 | 29

Attendance Monitoring System Through RFID and Face Detection

Chapter 3
LITERATURE REVIEW

3.1. Face Recognition And Detection

3.1.1. Real-Time Face Recognition System (RTFRS)

Suad Haji | Asaf Varol

Methodology:

They first collected the images of the faces. Then, they made sure the pictures were of good quality.
They then trained the system how to recognize faces in the pictures. They basically taught it some
rules or patterns to recognize the faces correctly. After a face was found, they used two methods called
Eigenfaces and Local Binary Patterns to pick out the important features of each face
They then trained the system to recognize faces using this info. They gave it many inputs of images to
train it for facial recognition. They checked how well the system works. They tested whether it could
recognize faces accurately in different situations and looked for ways to make it better

Conclusion:

In conclusion, the researchers have managed to create a Real-Time Face Recognition System
that employs a combination of Eigenfaces and LBP to achieve over 97% accuracy under
controlled conditions. In addition to the high performance, the developed system has real-
time processing capabilities. The joint use of the Eigenfaces and LBP features allows the
RFFRS to compensate for lighting differences and efficient scanning of large databases.
However, while the results under controlled conditions are promising, the system’s efficiency
falls significantly when recognizing faces captured at a random angle or distance. Still, the
conducted research proves that the proposed method of recognition has a significant potential
for application in real-world security systems. Further developments should focus on the
mentioned limitations to ensure the efficient recognition of faces under multiple conditions.
Dept of CSE Jan – May P a g e 3 | 29
Attendance Monitoring System Through RFID and Face Detection

3.1.2. Real-Time Face Detection and Segmentation

Alhussain AKOUM

Methodology:

The research focused on understanding the Viola-Jones algorithm a strong tool for spotting
faces quickly. This algorithm divides face recognition into three main steps. In the first step
the algorithm uses Haar-like features which are like small patterns resembling basic facial
features such as edges lines and rectangles. To speed up calculations the system uses
something called an integral image. This is a processed version of the original image that
makes it easy to calculate Haar-like features in any part of the image. The second step is
about picking the most useful features from the initial set. Here AdaBoost a machine
learning technique plays a role. AdaBoost acts like a filter carefully choosing the most
effective Haar-like features for telling faces apart from non-faces. It does this by training a
series of simple decision trees called weak classifiers based on individual features. Then it
combines these weak classifiers into a strong one giving priority to the most important
features for accurate face detection. Lastly to make things faster the Viola-Jones algorithm
uses a multi-layered structure called an attentional cascade. Images go through a series of
increasingly complex classification steps. Parts of the image that probably don t have a face
are quickly discarded saving time and resources. Only areas that seem promising go through
all the steps ensuring that faces are detected efficiently in the image.

Conclusion:

To sum up the research looked into how well the Viola-Jones algorithm can spot faces in
real-time. This approach with three steps breaks down face recognition for quicker
processing. The first step uses Haar-like features acting as templates for facial details along
with an integral image for quick calculations in the image. Then AdaBoost a machine
learning technique carefully selects the most useful features for distinguishing faces. It
trains simple decision trees based on individual features and combines them into a strong one
focusing on the most important features for accurate detection. Lastly to speed things up a
multi-layered attentional cascade is used. Images move through steps of increasing
complexity with areas unlikely to have faces discarded early on. This ensures efficient
detection by focusing on promising areas. The study s contribution lies in comparing a
standard Viola-Jones setup with a tweaked version that includes pre-processing steps. This
test suggests that pre-processing can improve detection accuracy. Overall the Viola-Jones
algorithm proves to be a strong and efficient solution for real-time face detection showing
potential for various applications needing fast and accurate face recognition.

Dept of CSE Jan – May P a g e 4 | 29

Attendance Monitoring System Through RFID and Face Detection

3.1.3. Frontal Face Generation Based Multi-angle Face Identification

System

Zihao Zhang | Huayan Zhang | Hui Liu | Shan Xin | Ning Xiao | Lei
Zhang

Methodology:

Modifying a face image taken from multiple angles so that it appears in front
view before applying facial recognition. facial features: the system starts by
detecting vital points on every video frame using a pre-trained model such as
Open Pose. In this context, the points are facial features that include the corners
of the eyes and mouth . Face rotation: for each frame, the system uses the
detected key points to calculate the face’s rotation angle required to place it in a
front view. This will enable the system to make necessary adjustments to ensure
that the face is fronted . Frontal face images creation using Generative
Adversarial Network : this step is accomplished using a specialized deep
learning model known as TP-GAN, which the system learns from a large
number of paired images where there is a frontal image and its corresponding
image from different angles. Understanding facial landmarks and rotation data,
TP-GAN generates a frontal oriented image . Facial recognition: when each
frame’s frontal face image is created, a standard facial recognition algorithm
such as ResNet-29 from dlib can be employed for identification . This system
has the ability to handle moving faces, unlike dlib, providing multiple angles of
the face.

Future Considerations:

Some of the potential limitations that have been proposed in the paper include;
Dataset Bias: Certain facial recognition systems trained on selected datasets
may exhibit bias. The authors recommend that TP-GAN should be trained on a
dataset specifically consisting of Asian faces to reduce the occurrence of this
bias . Other Challenges: Any additional conditions such as variations in light
and obstruction such as spectacles or hats should be handled, and additional
research is necessary to achieve a stronger model. Finally, the computation
Dept of CSE Jan – May P a g e 5 | 29
Attendance Monitoring System Through RFID and Face Detection

resource costs to process this system with deep learning forms like TP-GAN are
additional when used in real-time.The study demonstrates a novel approach to
improving the exploitation angle for efficient facial recognition in real-life
scenarios. The implications can assist various security, identification, and other
places.
3.1.4. Design of Face Detection and Recognition System for Smart
Home Security Application

Dwi Ana Ratna Wati | Dika Abadianto

Methodology:

The two steps in the process are : Face detection and recognition. Firstly a webcam captures an image.
This image is then checked by the face detection algorithm which looks for a match to a stored
template in the system. If a good match is found (indicating a face is present) the system proceeds to
recognition. For face recognition a method called Principal Component Analysis (PCA) is used. PCA
compares the image of the detected face with a database of authorized faces. The system calculates a
special distance measure (Euclidean distance) between the captured face and each face in the
database. The face in the database with the smallest distance is considered a potential match. However
to avoid incorrect identifications this distance must be below a set threshold. If it meets this condition
the system recognizes the person and grants access.

Conclusion:

The study made a facial recognition system for smart home security using specific tools. It worked
okay, getting about 80% right in ideal conditions. But it had trouble when faces were partly covered,
like with hats, or if someone wore clothes with similar colors to their face.

To make it better in real-life situations, they think it's a good idea to add passwords along with face
recognition. In the future, they could make the system stronger by training it on a wider range of faces
and dealing with different lighting situations.

3.1.5. Multiple Face Detection Attendance System

Dept of CSE Jan – May P a g e 6 | 29

Attendance Monitoring System Through RFID and Face Detection

Raj Kaste | Harish Pandilla | Priyesh Surve | Mubin Shaikh | Shalaka

Deore

Methodology:

The system accomplishes multi-face detection and recognition through two stages. Initially
it detects faces in an image. Here input images are turned into grayscale for efficiency and
then a Haar cascade classifier is employed. This classifier is adept at identifying faces by
pinpointing specific features like contrasting rectangular areas using the integral image
technique for efficient calculation. By combining multiple weak classifiers it achieves high
accuracy in detecting faces which are then outlined with boxes in the image. The second
stage concentrates on recognizing the detected faces. In a training phase the system creates
Eigenfaces capturing key facial variations from a set of training images. To simplify
processing Principal Component Analysis (PCA) is used to reduce data complexity while
retaining important facial features. During recognition PCA is directly applied to new
images (containing faces to be recognized) to extract their features. Finally the Euclidean
distance is used to compare the features of a new face with those stored in the training data
(represented by Eigenfaces). The face in the training data with the most similar features
(smallest Euclidean distance) is recognized.

Conclusion:
In summary, this research paper presents a new method of taking attendance using face recognition. It
involves getting faces and recognizing them. For the first purpose, the system uses a tool called Haar
cascade classifier to detect faces imprecisely. Afterwards, Eigenfaces in combination with Principal
Component Analysis recognizes the face by concentrating on the significant attributes kept within the
system .There is a possibility of how this system can be used on computers for exam entry, and it is
probable that this research will first be utilized in actual situations where people must manage
attendance securely and effortlessly. It can be improved by facilitating recognition from distinct
angles and making the program compatible with smartphones. To sum up, this research offers an
efficient and robust system for automatically detecting presence through facial recognition.

3.1.6. Multi-Faces Recognition Process Using Haar Cascades and

Eigenface Methods

Teddy Mantoro | Suhendi

Methodology:
Dept of CSE Jan – May P a g e 7 | 29
Attendance Monitoring System Through RFID and Face Detection

The proposed method addresses multi-face recognition through two main steps: detection and
recognition. Initially grayscale images are generated by converting the original ones and
reducing their color depth. Then robust classifiers known as Haar cascades are employed to
detect faces. These cascades identify specific features in the image such as contrasting
rectangular areas using the integral image technique for efficient calculation. By combining
weak classifiers strategically they achieve high accuracy in face detection with detected
faces marked using bounding boxes. Recognition relies on feature extraction. During training
Eigenfaces capturing key variations in faces are generated from the training data using
Principal Component Analysis (PCA) for dimensionality reduction. PCA helps focus on
important features while managing data complexity. During recognition PCA is directly
applied to test images to extract their features. Finally the Euclidean distance a measure of
similarity is calculated between the features of a test face and those in the training data
determining the closest match as the recognized face.

This two-step process implemented using a webcam proves particularly useful for a
computer-based exam registration system. During registration captured images undergo face
detection and recognition associating individuals with the exam database. On exam day the
system authenticates participants by recognizing their faces before granting access to the
exam ensuring only registered individuals take the test.

Conclusion:

In conclusion, this paper presents a good way of recognizing multiple faces using both Haar
cascades and Eigenfaces. This is very accurate since it can get over 91% correct even when
the faces are at a distance of 2 meters. This method is excellent because it is faster compared
to normal recognizing a single face each time, which is beneficial when it comes to multiple
people identification fast. It was even practical with a computer-based examination system,
which means this is beneficial in real work where security is a priority. To enhance this even
more, they recommend this work can be improved on the ability to recognize faces at various
angles and perhaps also make it available in phones. In conclusion, this research provides an
efficient, strong way of recognizing multiple faces, and this is very beneficial in the face
recognition world.

3.2. RFID TECHNOLOGY

3.2.1. THE RFID TECHNOLOGY AND ITS APPLICATIONS:
A REVIEW

DAVINDER PARKASH | TWINKLE KUNDU | PREET KAUR

Dept of CSE Jan – May P a g e 8 | 29

Attendance Monitoring System Through RFID and Face Detection

RFID stands for Radio Frequency Identification and is essentially a more clever barcode.
Instead of requiring a clear line of sight to the tag, it uses radio waves to scan things from a
long distance. Its components are tags, antennas, the reader, communication infrastructure,
application software. The tag is a clever sticker with a chip and an antenna, storing data
Wireless to a reader by request. In contrast, the antenna interacts with the tags and both sends
and receives radio wave transmission, a reader matches with an antenna and sends
information to a processing unit and likely a computer system . The communication
infrastructure enables the reader to communicate with the central database and/or software.
This software helps you organize all the information gathered via the RFID system. RFID has
several advantages since it does its job rapidly and correctly, over many tags to be read at
once, read at a long distance or unlimited line of sight, it can handle more data than barcodes
and a few tags can be rewritten. It is also limited by costs, range, some materials disrupt the
signal, putting chips on endangers the purpose of privacy. In the future, it could become a
more efficient and cheaper option, more and more business could start using it, RFID could
be an excellent combination with sensors for accurate live data on a lot of stuff. Making new
standards can help make RFID systems work with each other. It could enhance the industry in
identifying and tracking.

3.2.2. RFID Technology and Its Applications in Internet of Things

(IOT)

Xiaolin Jia | Quanyuan Feng | Taihua Fan | Quanshui Lei

Introduction to RFID Technology and IoT:

RFID (Radio Frequency Identification) is an automatic technology that helps identify objects
record metadata or control individual targets through radio waves. It is considered essential
for the Internet of Things (IoT) where RFID readers connected to the internet can identify
track and monitor tagged objects globally and in real-time.

Components of RFID System:

• It includes tags (transmitters/responders) and readers (transmitters/receivers).

• Tags have a microchip and antenna while readers communicate with tags using radio
waves.
• RFID enables automated identification and data capture bringing significant changes
to business activities.

Layers of IoT Architecture:

Dept of CSE Jan – May P a g e 9 | 29

Attendance Monitoring System Through RFID and Face Detection

• Perception Layer: Gathers information from the physical world using sensors RFID
systems cameras etc.
• Network Layer: Offers transparent data transmission capability through various
communication networks.
• Service Layer: Comprises data management and application service sub-layers
aiding data processing and providing user interfaces.

Applications of RFID Technology:

Various fields benefit from RFID technology including logistics manufacturing

agriculture healthcare transportation and retailing. RFID aids in supply chain
management production process control object tracking and monitoring. Specific
applications range from antitheft vehicle immobilizers to assembly line automation in the
automotive industry.

Challenges of RFID Technology:

• Collision Problems: Simultaneous transmissions lead to collisions requiring efficient

anti-collision protocols.
• Security and Privacy Concerns: Vulnerabilities to eavesdropping spoofing and
denial of service attacks pose risks to organizations and individuals.
• Cost Design and Integration: RFID tags are still more expensive than printed labels
require reliable engineering and need effective integration into existing systems.

Conclusion:

IoT relies on technologies like RFID to create extensive networks for information sensing
and intelligent processing. Despite challenges RFID offers vast benefits and ongoing
research aims to address security cost and integration issues. The evolution of IoT aligns
with advancements in RFID and sensor technologies leading to innovative applications and
transformative changes in various industries.

Dept of CSE Jan – May P a g e 10 | 29

Attendance Monitoring System Through RFID and Face Detection

3.2.3. RFID Based Attendance System

Ajay Joshi | Aman Ahmad | Arpit Saxena | Poonam Juneja

Introduction of RFID Technology

RFID, Radio Frequency Identification, is a system with a card and a reader linked by
wireless. It is a current solution to attendance management in schools and supports the digital
environment. Initial tests demonstrate it has simplified the process of taking attendance.

Methodology

Hardware

• RFID reader – fetches data from RFID tags without physical contact.
• RFID tags – contain an antenna and a chip used for storing information such as an ID.
• Arduino UNO – a board used for processing, storing, and utilizing data.
• DS 3231 RTC module – utilized for maintaining time.
• Liquid crystal display – used for showing time and the status of attendance forms.

Software

The software used in coding:

• Arduino IDE.
• Proteus Design Suite – helps in simulating circuits.

Operation Principle

Tags are supplied to individuals, which communicate with the reader to forward their data to
the microcontroller.

The microcontroller approves the storage of the attendance in a database and confirms
whether the student has attended the class.

It uses an LCD display to show information about the student and an RTC module to
maintain time.

Future Plans

Dept of CSE Jan – May P a g e 11 | 29

Attendance Monitoring System Through RFID and Face Detection

In places where contactless can be critical, RFID may perform extra tasks.

Add in additional features like a fingerprint scanner.

It will also not only be limited to attendance but will also be used in other places like library
management, parking and safety, etc.

Conclusion

This system is inexpensive, accurate, and straightforward in its operation in the collection of
attendance.

It allows the majority of schools to take a step in the future of going digital and contact-less,
making the functionality very affordable.

It can easily be implemented whenever need be.

Chapter 4
DATA
Our Automated Attendance System involves Real-Time Face Recognition and Radio
Frequency Identification , which works more effectively than the traditional approach. The
following steps will help make it work: Collecting Real-World Data – the taking of good
pictures of students’ faces under well-controlled lighting Table and linking it to their ID with
an RFID tag or something secure for their privacy . However, we could also use simulation to
make it work more accurately: Simulate Data – we mean watching how everything goes
down to make our system more outstanding , and if glasses are thrown into play, they’re
thrown into play, and we give an explanation. Some of the other calculations will be based on
some In addition to providing the simulation of real cases to train the RTFR engine to
function under different circumstances, No unauthorized person can enter the building if an
RFID tag is not present. Our Automated Attendance System will provide greater accuracy
and efficiency by using both real-world data and simulation techniques.

Dept of CSE Jan – May P a g e 12 | 29

Attendance Monitoring System Through RFID and Face Detection

Chapter 5
SYSTEM REQUIREMENTS SPECIFICATIONS

Automated Attendance Management System RTFR and RFID Software Requirements:

System Overview: Attendance marking system for students with automation Increase
accuracy and security with RTFR and RFID Secure information User-friendly

User Roles: Administrator: Manages user profiles, settings, and permissions Registered User:
Allows to register your face, marks attendance using RFID, attends and holds the face
scanner, and views attendance details

Functional Requirements: User Management RFID Tag Management Record the facial data

Real-Time Face Recognition Attendance Marking Reporting

Non-Functional Requirements: Performance: Responds within 2 seconds Security: Follows

data privacy laws Usability: Documentation available and user-friendly Reliability: Disaster
recovery Maintainability: Modular, maintainable code Scalability 3. Lighting conditions and
user consent.

Dept of CSE Jan – May P a g e 13 | 29

Attendance Monitoring System Through RFID and Face Detection

Chapter 6
SYSTEM DESIGN

6.1 Server:
• Manages user accounts (students faculty).

• Saves student information (ID name department year and facial picture hash).

• RFID tag IDs are associated with student profiles.

• Logs attendance data (date time student ID course ID and attendance status).
Dept of CSE Jan – May P a g e 14 | 29
Attendance Monitoring System Through RFID and Face Detection

• RTFR for facial recognition engine

Communication to the mobile app or the client application for data exchange

Database:

A secured database management system records user information, RFID tag relationship, and
attendance information.

RFID Scanner:

Installed RFID readers at the classroom entrance or other relevant stations which have tags of with a
validated format

Read students’ RFID tags

Raspberry Pi Camera:

A camera placed at a strategic point adjacent to the RFID scanner which has sufficient resolution and
frame rate for accurate facial recognition

Capture the image of the student’s face after successful RFID read

Dept of CSE Jan – May P a g e 15 | 29

Attendance Monitoring System Through RFID and Face Detection

Chapter 7
IMPLEMENTATION AND PSUEDOCODE

# Initialize Functions
FUNCTION CaptureFaceImage(camera)
# Capture image from camera
# Return captured image
END FUNCTION

FUNCTION RecognizeFace(image, known_faces)

# Use face recognition algorithm (e.g., LBPH) to identify face in image
# Compare against known faces database
# IF match found
# RETURN name of recognized person
# ELSE
# RETURN "Unknown"
# END IF
END FUNCTION

FUNCTION VerifyRFID(reader)
# Read ID from RFID reader
# Check ID against authorized users database
# IF valid ID found
# RETURN user ID
# ELSE
# RETURN "Invalid ID"
# END IF

Dept of CSE Jan – May P a g e 16 | 29

Attendance Monitoring System Through RFID and Face Detection

END FUNCTION

# Main Loop
LOOP
# Read ID from RFID reader
user_id = VerifyRFID(reader)

# Check for valid ID

IF user_id == "Invalid ID"
# Display error message (optional)
ELSE
# Capture face image
image = CaptureFaceImage(camera)

# Recognize face in captured image

recognized_name = RecognizeFace(image, known_faces)

# Check if recognized face matches RFID user

IF recognized_name == user_id
# Attendance confirmed, log attendance (store user ID and timestamp)
# Display success message (optional)
ELSE
# Attendance mismatch, log event (optional)
# Display warning message (optional)
END IF
END IF

# Wait for next user

Dept of CSE Jan – May P a g e 17 | 29

Attendance Monitoring System Through RFID and Face Detection

END LOOP

Functions: We define functions for capturing face images, recognizing

faces, and verifying RFID tags.
Main Loop: The system continuously loops, checking for RFID tags
and capturing faces.
RFID Verification: The VerifyRFID function reads the ID from the
reader and checks it against a database of authorized users.
Face Recognition: If a valid ID is found, the CaptureFaceImage
function captures an image, and the RecognizeFace function attempts
to identify the face using the chosen algorithm (e.g., LBPH) against a
database of known faces.
Matching: The system checks if the recognized name matches the
RFID user ID.
Attendance Confirmation: If there's a match, attendance is confirmed
and logged.
Mismatch Handling: If there's a mismatch, the event is logged
(optional), and a warning message can be displayed.
Loop continues: The system waits for the next user to come forward.
Chapter 8
CONCLUSION OF CAPSTONE PROJECT PHASE – 1

This report outlines the achievements of Phase 1 in developing an Automated Attendance

Management System using Real-Time Face Recognition (RTFR) and Radio Frequency
Identification (RFID) technologies.

8.1. Key Accomplishments:

Dept of CSE Jan – May P a g e 18 | 29

Attendance Monitoring System Through RFID and Face Detection

8.1.1. Literature Review. The research of 15 scientific articles of RTFR, RFID integration,
and attendance management systems. The review allowed developing our entire system’s
concept.

8.1.2. System Architecture. The high-level description of the developed system was based on
the existing systems and research. The interactions of hardware, RFID reader, camera, and
software- facial recognition engine, attendance database, were outlined to promote
orderliness during development.

8.1.3 Hardware Component. Based on specifications, costs, and compatibility assessments,

the key hardware components were chosen: an RFID reader with the requisite tag format and
a Raspberry Pi camera for facial recognition. The selected components positively affect the
system’s performance.

8.2 Conclusion. The Phase 1 of developing our attendance management system is complete.
The research, teleology’s definition, and hardware selection is ready for the further steps.
These are occurring the software, connecting and integration of hardware, and system testing
for performance and accuracy.

Chapter 9
PLAN OF WORK FOR CAPSTONE PROJECT PHASE – 2

Building on the foundation set in Phase 1 Phase 2 of the Automated Attendance System
project focuses on making significant strides towards a fully operational system. Here is a
breakdown of our main objectives and goals:

9.1 Refining the System Design

Review the current system design as presented in Phase 1.

Identify areas for improvement; priority, security, scalability, and user experience.

Incorporate best practices from recent research in Real-Time Face Recognition and Research
on RFID based attendance system

9.2 Software development

Develop user registration and management for administrators and students

Development of RFID tag management to link tags with users profile

Dept of CSE Jan – May P a g e 19 | 29

Attendance Monitoring System Through RFID and Face Detection

Real-time face recognition from your chosen RTF library or framework

Attendance data logging and reporting

9.3 System Integration and testing

Integrate software modules with hardware components present the RFID reader, camera for
smooth communication and exchange of data.

Create a testing roadmap with several scenarios

Successful RFID tag read, and facial recognition 2. Handling of errors invalid tags, and
unrecognizable faces.

Performance evaluation within different lighting conditions

Conduct security test to ensure data integrity and Unauthorized access.

9.4 Deployment

Develop a deployment scale taking into consideration; network connectivity, a user’s access
control system, and data security.

Develop a user training session Plausible and rollout plan at the preferred testing environment
for the final deployment.

Phase 2 of the Automated Attendance System will have a functional prototype that is
operationally tested and successfully rolled out in the chosen simulated environment.

References/Bibliography

• Kushal M Kushal Kumar B V Charan Kumar M J Prof.Pappa M “ID Card

Detection with Facial Recognition using Tensorflow and OpenCV” Proceedings of the
Second International Conference on Inventive Research in Computing Applications
(ICIRCA-2020) IEEE Xplore Part Number: CFP20N67-ART; ISBN: 978-1-7281-
5374-2
• Zihao Zhang Huayan Zhang Hui Liu Shan Xin Ning Xiao Lei Zhang “
Frontal Face Generation Based Multi-angle Face Identification System“2021
International Conference on Computer Control and Robotics

Dept of CSE Jan – May P a g e 20 | 29

Attendance Monitoring System Through RFID and Face Detection

• Alhussain AKOUM “Real Time Face Detection and Segmentation” International

Journal of Applied Engineering Research ISSN 0973-4562 Volume 13 Number 19
(2018) pp. 14308-14312 © Research India Publications. http://www.ripublication.com
• DAVINDER PARKASH TWINKLE KUNDU & PREET KAUR “THE RFID
TECHNOLOGY AND ITS APPLICATIONS: A REVIEW”
• Dwi Ana Ratna Wati Dika Abadianto “Dwi Ana Ratna Wati Department of Electrical
Engineering Universitas Islam Indonesia Yogyakarta Indonesia [email protected]
Dika Abadianto” 2017 2nd International Conferences on Information Technology
Information Systems and Electrical Engineering (ICITISEE)

Dept of CSE Jan – May P a g e 21 | 29