© January 2024| IJIRT | Volume 10 Issue 8 | ISSN: 2349-6002

Implementation of Face Recognition in Attendance

System

Mrs. Nethra H L1, Shashank Dubey2, Shatrunjay Kumar3, Sridhar Srinath4, Sudarshan Singh5
1,2,3,4,5
Department of Computer Science & Engineering, Dayananda Sagar Academy of Technology and
Management (DSATM), Bengaluru

Abstract— The Attendance Management System based on entertainment, civil economy, and public safety
Face Recognition (FRAMS) represents a groundbreaking domains[1]. It has become essential for general
approach crafted to simplify and automate the conventional organizations to monitor their employees' attendance
attendance tracking procedures across diverse educational
systematically. However, errors are commonly
and corporate environments. Utilizing cutting-edge facial
introduced during the formulation of these attendance
recognition technology, FRAMS offers a secure, effective,
and user-friendly approach to precisely record and manage systems. An examination of the existing fingerprint-
attendance data. based attendance system revealed situations where
Employing advanced deep learning algorithms, the system fingerprints might not register accurately, resulting in an
engages in the identification and authentication of error rate of approximately 5%.
individuals based on their facial features. Through the The design of the face recognition-based attendance
utilization of a repository containing pre-registered faces, system has a positive impact on both current and future
the system guarantees a high degree of accuracy and enterprise development. This essay aims to create a real-
reliability in attendance tracking, obviating the necessity
time image processing-based facial recognition time and
for manual recording techniques like paper-based systems
attendance system.[2] Through real-time image
or swipe cards.
processing, four investigative experiments were
Index Terms— Image processing, face recognition conducted: evaluating the accuracy rate during actual
technology, face recognition, attendance system. check-in, assessing the stability of the face recognition
time and attendance system, analyzing the skip rate, and
I. INTRODUCTION examining the interface settings.
The experimental findings illustrate that the integration
The era of Internet growth has seen significant impacts of face recognition technology and computer assistance
from computer technology on various aspects of people's in the time and attendance system effectively achieves
lives and careers. There is a steady increase in instances the desired outcomes, reflecting the overall feasibility of
where individuals interact with computers and a the algorithm's design. Students who successfully
corresponding rise in computing usage. The field's completed the attendance check-in method did so
projects, marked by immense inventiveness, hold vast promptly, eliminating the need for a complicated roll call
potential applications, making it a challenging subject. and promptly grasping its purpose. The transition to the
Facial recognition plays a crucial role as a distinguishing new attendance system and the future system time has
mark for individuals, influenced by the progressive seen significant progress, enhancing the reliability of face
impact of technology on people's lives. The association recognition technology and improving attendance rates.
of artificial intelligence and recognition in computing A more comprehensive investigation and implementation
devices presents a formidable innovation with numerous of these advancements are recommended for further
potential applications, establishing it as a particularly development. Significant progress has been made in
challenging subject in the field. enhancing the dependability of face recognition
In recent years, there has been a significant surge in the technology and the attendance rate through notable
popularity of facial recognition application systems, improvements in the transition form for the attendance
especially as a tool for computer security. Face system and future system time. Our scientists should
recognition technologies find common uses in home

IJIRT 162129 INTERNATIONAL JOURNAL OF INNOVATIVE RESEARCH IN TECHNOLOGY 75

 © January 2024| IJIRT | Volume 10 Issue 8 | ISSN: 2349-6002

thoroughly explore and implement these advancements and the format of attendance system conversion have
for a more comprehensive understanding. significantly enhanced the attendance rate and the
In recent times, the swift development of face recognition reliability of face recognition technology. This progress
applications as a computer security technology has merits further exploration and implementation by our
garnered substantial attention globally, particularly in the scientists.
current climate of rampant terrorist activities. This II. HISTORY
technology finds numerous applications in public safety,
civil economy, and home entertainment. The recording of Several researchers have already emphasized on the
personnel attendance has become a fundamental necessity of automating classroom attendance, asserting
requirement for general enterprises. However, the that manual methods are time-consuming. Despite the
formulation of attendance systems often introduces availability of automatic options like biometric
unnecessary errors. For instance, the current fingerprint attendance, these methods also cause time wastage as
attendance system exhibits an error rate of approximately students must line up to imprint their thumb on scanning
5%, With instances where fingerprints fail to register, devices. This study introduces an efficient algorithm for
significantly impeding attendance efficiency, especially automated attendance marking. Utilizing a camera placed
in large attendance settings prone to congestion. On the in the classroom, it continuously takes images of
other hand, card attendance systems are susceptible to students, detects faces, compares them with a database,
employees swiping cards for others, hindering real-time and records attendance. The paper conducts a review of
attendance goals. In comparison, the face recognition related work in attendance systems, talking about the
system boasts higher accuracy and stability due to the system architecture, algorithm used, and results found.
multitude of recognition points, surpassing other systems Erik Hjelmas' face detection plays a pivotal role as an
and mitigating congestion issues. Despite China's late initial and robust step, with its primary objective being
start in face recognition research, scientific researchers the identification, localization, and extraction of the
have rapidly advanced, establishing significant industry facial region from the background. This procedure is
positions in the field. With the advent of the big data era utilized in various applications, including content-based
and the commercial viability of face recognition image retrieval, video coding, video conferencing, crowd
technology, the future of this research appears promising, surveillance, and the development of intelligent human-
with substantial market demand. computer interfaces. Despite recent attention, the face
The development of enterprises is positively influenced detection problem has only recently gained substantial
by the design of a face recognition attendance system focus among researchers. The human face, being a
incorporating real-time video processing. This article dynamic object with a high degree of variability, poses
aims to create a face recognition time and attendance challenges in computer vision. Various techniques,
system based on this real-time image processing. Three ranging from simple edge-based algorithms to
investigative experiments were conducted in this sophisticated high-level approaches employing advanced
experiment: assessing the accuracy rate of the face pattern recognition methods, have been proposed to
recognition system during actual check-in, evaluating the address this complexity.
stability of the face recognition time and attendance
system employing real-time image processing, and PROPOSED SYSTEM
configuring the interface settings of the face recognition REAL-TIME VIDEO FACE IMAGE RECOGNITION
attendance system using real-time image processing. The FACIAL IDENTIFICATION
experimental findings confirm that the time and At the heart of the entire identification process lies facial
attendance system, with the integration of face recognition, a computer vision technology designed for
recognition technology and computer assistance, attains identity authentication by analyzing facial feature
the anticipated results, fully demonstrating the feasibility information. In a broader context, facial recognition
of the overall algorithm's design. Students who utilized encompasses two key components: face detection and
the attendance sign-in system efficiently completed their facial feature matching. The technology relies on the
tasks, eliminating the need for the cumbersome roll call, facial features of individuals and processes input face
swiftly understanding the system's operation and images or video streams. Facial recognition, falling under
functionality. The innovations in the future system time the category of biometric recognition technology,

IJIRT 162129 INTERNATIONAL JOURNAL OF INNOVATIVE RESEARCH IN TECHNOLOGY 76

 © January 2024| IJIRT | Volume 10 Issue 8 | ISSN: 2349-6002

comprises four essential components: face image contrast, a negative result at any juncture leads to the
collection, face image preprocessing, face image feature instantaneous dismissal of the sub-window
extraction, and matching, combined with hardware The cascade training process entails navigating two
cameras, network lines, and computing devices. The trade-offs. Frequently, classifiers equipped with more
calculation method is as follows. features tend to achieve elevated detection rates while
T = min {T1, T2, …, Tn} simultaneously minimizing false positive rates.
The underlying principle of facial recognition involves However, classifiers with more features also demand
biometric technology that acquires facial information more computation time. At its core, the trade-offs
through camera equipment and preprocesses it encompass considerations such as the quantity of
accordingly. The initial phase, face detection, serves to classifier stages, the quantity of features within each
determine the presence of a human face image in the stage, and the threshold set for each stage, all aimed at
overall image, establishing its size, position, and minimizing the anticipated number of assessed features.
segmenting the detected human face image into the facial
region. The final stage is facial recognition, where facial
feature information and image details are extracted to
confirm its presence in the repository. If a match is found,
corresponding identity information is established;
otherwise, no recognition results are obtained.

III. METHODOLOGY

In the operation of this system, a two-step mechanism is Fig 1: Cascade classifier

employed. Initially, the process involves face detection, the number of classifier stages, ii) the number of features
followed by face recognition. Viola-Jones face detection in each stage, and iii) the threshold of each stage, are
algorithm is utilized for the face detection stage, while traded off in order to minimize the expected number of
the face recognition stage employs a hybrid algorithm evaluated features.
combining PCA (Principal Component Analysis) and Regrettably, identifying this optimum presents a
LDA (Linear Discriminant Analysis). significantly challenging problem. In practical terms, a
straightforward framework is employed to create an
VIOLA- JONES ALGORITHM
efficient classifier that proves highly effective. Every
The Viola-Jones algorithm consists of three primary
stage within the cascade plays a role in reducing the false
components: Integral Images, Ada-Boost Algorithm, and
positive rate while simultaneously lowering the detection
Attentional Cascade. The integral image computes a
rate. This process includes establishing a target for the
value for every pixel (x, y) by summing the pixel values
minimum reduction in false positives and the maximum
located above and to the left of the position (x, y). This
decrease in detection. The training of each stage involves
computation occurs swiftly in a single pass through the
the incremental addition of features until the predefined
image. The Viola-Jones algorithm utilizes Haar-like
target rates for both detection and false positives are met.
features, representing the scalar product between the
This determination is made by evaluating the detector on
image and certain Haar-like structures. Feature selection
a validation set. Further stages are integrated until the
is accomplished through Ada-Boost, which provides an
overall goal for false positives and detection rate is
effective learning algorithm and robust bounds on
reached.
generalization performance.
IV. WORKING OF SYSTEM
FLOW CHART
Step1: Commencement: The process initiates with the
The overall structure of the detection process resembles
acquisition of the input image from the device.
a degenerate decision tree known as a "cascade." Upon a
favorable outcome from the initial classifier, the Step 2: Facial Detection:
assessment proceeds to a second classifier, tailored to Utilization of Haar classifiers is integral to the detection
optimize detection rates. Subsequent positive results process. Initially, the face detection algorithm undergoes
trigger additional classifiers in a cascade fashion. In testing with a diverse set of images, encompassing

IJIRT 162129 INTERNATIONAL JOURNAL OF INNOVATIVE RESEARCH IN TECHNOLOGY 77

 © January 2024| IJIRT | Volume 10 Issue 8 | ISSN: 2349-6002

variations in face positions and lighting conditions. V. FUTURE SCOPE

Following this, the algorithm is deployed to enable real-
time face detection in videos. The training of the The future scope of the attendance system utilizing face
algorithm involves using facial images, and its recognition technology holds immense potential for
application extends to classroom images for the detection further advancement and integration into diverse
of multiple faces within the frame. domains. One significant avenue is the exploration of
enhanced security measures. By incorporating advanced
Step 3: Face Recognition and Attendance: algorithms and biometric features, the face recognition
After the completion of the face detection phase, the attendance system can contribute to robust security
subsequent step entails face recognition. This is achieved protocols in sensitive environments such as government
by cropping the initially identified face from the image institutions, financial organizations, and research
and subsequently comparing it with the database—a facilities.
process commonly referred to as region-of-interest Moreover, the scalability and adaptability of the face
selection. Subsequently, individual verification of recognition attendance system offer opportunities for
students' faces is carried out against the face database widespread adoption in various educational and
using the Eigen Face method, leading to attendance corporate settings. The system's ability to efficiently
marking on the server. manage large datasets and accurately record attendance
makes it a viable solution for institutions with diverse
Step 4: Student Verification: attendance tracking needs. In the future, we can
The system integrates a camera to capture classroom anticipate the integration of artificial intelligence (AI) to
images, transmitting them to the image enhancement enhance the system's learning capabilities, leading to
module. After the enhancement process, the image improved accuracy and adaptability in different
progresses to the Face Detection and Recognition scenarios.
modules, culminating in the marking of attendance on the The increasing prevalence of remote work and online
database server. In the enrollment phase, templates of education presents another avenue for the future
students' facial images are stored in the Face database. expansion of face recognition attendance systems. As
The algorithm detects all faces from the input image, virtual and hybrid learning environments become more
sequentially comparing them with the face database. common, the integration of face recognition technology
Upon recognizing a face, attendance is recorded on the can play a pivotal role in monitoring attendance and
server, making it accessible for various purposes. The engagement during virtual sessions. Additionally, the
system adheres to a specific protocol for attendance. system could be further optimized for seamless
integration with existing online platforms, providing a
comprehensive solution for institutions adapting to the
evolving landscape of education and work.
Furthermore, the continuous development of hardware
technology, particularly in camera and processing
capabilities, will likely contribute to the refinement of
face recognition algorithms. Higher-resolution cameras
and faster processors can lead to more accurate and faster
recognition, improving the overall efficiency of the
attendance system. Collaborations between technology
developers and educational institutions or businesses
could facilitate the integration of cutting-edge features
and ensure the system remains at the forefront of
technological advancements.
In conclusion, the future scope of the attendance system
using face recognition extends beyond its current
Fig 2: Flowchart of the System applications. The ongoing advancements in technology,
coupled with the system's adaptability and scalability,

IJIRT 162129 INTERNATIONAL JOURNAL OF INNOVATIVE RESEARCH IN TECHNOLOGY 78

 © January 2024| IJIRT | Volume 10 Issue 8 | ISSN: 2349-6002

position it as a versatile solution with potential In summary, the face recognition attendance system
applications in security, education, remote work, and represents a transformative solution with broad
beyond. As research and development in this field applications. Its successful implementation in
progress, we can anticipate a more sophisticated and educational and corporate contexts underscores its
widespread implementation of face recognition potential to redefine traditional attendance management
attendance systems, contributing to enhanced efficiency practices. As technology continues to evolve, further
and security in various sectors. enhancements and widespread adoption can be
anticipated, solidifying the face recognition attendance
VI. CONCLUSION system as a cornerstone in the evolution of attendance
tracking methodologies.
In conclusion, the implementation of an attendance The attendance system realizes the expected attendance
system using face recognition technology marks a results through face recognition technology with the help
significant leap forward in streamlining attendance of a computer, which fully rejects the feasibility design
tracking processes across diverse sectors. The system's of the overall algorithm. The students who have
success lies in its ability to combine cutting-edge facial completed the attendance sign-in system quickly
recognition algorithms with real-time video processing, completed the tasks, got rid of the complicated sign of
creating a seamless and efficient method for recording roll call, and soon realized the sign of operation and
attendance. As demonstrated through various function. The system has made tremendous innovations
experiments and applications, the face recognition and greatly improving the attendance rate.
attendance system not only enhances accuracy but also
mitigates the shortcomings associated with traditional REFERENCES
methods, such as manual recording or card-based
systems. [1] Yang, Hao, and Xiaofeng Han. "Face recognition
One notable advantage of this technology is its potential attendance system based on real-time video
to revolutionize attendance management in educational processing." IEEE Access 8 (2020): 159143-
institutions. By automating the attendance tracking 159150.
process, educators and administrators can allocate more [2] Mustakim, Nafis, et al. "Face recognition system
time and resources to focus on core educational activities. based on raspberry Pi platform." 2019 1st
The system's adaptability to different lighting conditions International Conference on Advances in Science,
and face orientations further contributes to its reliability, Engineering and Robotics Technology (ICASERT).
ensuring consistent performance in various IEEE, 2019.
environments. [3] Elias, Shamsul J., et al. "Face recognition attendance
Beyond education, the face recognition attendance system using Local Binary Pattern (LBP)." Bulletin
system shows promise in corporate settings, where of Electrical Engineering and Informatics 8.1
accurate attendance records are crucial for payroll (2019): 239-245.
management and workforce analysis. The integration of [4] Bah, Serign Modou, and Fang Ming. "An improved
such systems aligns with the evolving landscape of face recognition algorithm and its application in
workplace technology, facilitating a more efficient and attendance management system." Array 5 (2020):
secure environment. 100014.
While acknowledging the achievements of the current [5] Mamatkulovich, Babakulov Bekzod.
face recognition attendance systems, it is essential to "AUTOMATIC STUDENT ATTENDANCE
consider ongoing research and development. Future SYSTEM USING FACE RECOGNITON." Next
advancements may introduce refinements to address Scientists Conferences. 2022.
potential challenges, improve recognition accuracy, and [6] Teoh, K. H., et al. "Face recognition and
enhance the overall user experience. As with any identification using deep learning approach." Journal
technology, continuous innovation remains pivotal to of Physics: Conference Series. Vol. 1755. No. 1. IOP
adapting the face recognition attendance system to Publishing, 2021.
emerging needs and ensuring its sustained effectiveness.

IJIRT 162129 INTERNATIONAL JOURNAL OF INNOVATIVE RESEARCH IN TECHNOLOGY 79

 © January 2024| IJIRT | Volume 10 Issue 8 | ISSN: 2349-6002

[7] Shetty, Anirudha B., and Jeevan Rebeiro. "Facial

recognition using Haar cascade and LBP classifiers."
Global Transitions Proceedings 2.2 (2021): 330-335.
[8] Hangaragi, Shivalila, Tripty Singh, and N. Neelima.
"Face detection and Recognition using Face Mesh
and deep neural network." Procedia Computer
Science 218 (2023): 741-749.
[9] Damanik, Rudolfo Rizki, et al. "An application of
viola jones method for face recognition for absence
process efficiency." Journal of Physics: Conference
Series. Vol. 1007. No. 1. IOP Publishing, 2018.

IJIRT 162129 INTERNATIONAL JOURNAL OF INNOVATIVE RESEARCH IN TECHNOLOGY 80