ISSN 2347–3657

Volume 13, Issue 1, Jan 2025

AUTOMATED ATTENDANCE SYSTEM USING FACE

RECOGNITION
1
A.Anitha,2Rohith Reddy Maddi,3Poojitha Nalavath,4Uday Babu Parise,5ch.Vinay Kumar
1
Assistant Professor,Department Of Computer Science and Engineering, Sreyas Institute of
Engineering and Technology, anitha.a@sreyas.ac.in.
2,3,4,5
Students,Department of Computer Science and Engineering, Sreyas Institute of Engineering and
Technology.
2
rohithreddymaddi28@gmail.com, 3poojithanalavath@gmail.com, 4udayparise123@gmail.com,
5
chitiyalavinay@gmail.com.

ABSTRACT
Attendance management is a vital daily activity in educational institutions,
workplaces, and organizations, traditionally carried out using manual methods such as
roll calls or name-based identification. These conventional approaches are time-
intensive, prone to human error, and inefficient, especially in large groups. This
project proposes an automated attendance system that leverages face recognition
technology to address these challenges and modernize the process. The system is
designed to be installed in classrooms or similar environments, where it captures and
processes the facial data of individuals for attendance tracking. Students' information,
including their name, roll number, class, section, and facial images, is pre-registered
and stored in a structured dataset. Using the OpenCV library, the system extracts,
processes, and trains the facial images to create a robust recognition model. Before
the start of a class, students interact with the system, which scans their faces, matches
the captured data with the stored dataset, and automatically records attendance upon
successful identification. By eliminating the need for manual intervention, the system
enhances accuracy, saves time, and ensures a seamless and efficient attendance
process. Furthermore, it aligns with the growing need for digital transformation by
introducing a secure, user-friendly, and technologically advanced solution that
improves operational efficiency and minimizes errors.
KEYWORDS: Attendance System, Face Recognition, Automated Process, Manual
Process, OpenCV, Student Information, Roll Number, Photographs, Time
Management, Classroom Device, Dataset Training, Modernization, Automation.

23
 ISSN 2347–3657
Volume 13, Issue 1, Jan 2025

I.INTRODUCTION
To verify the student attendance record, Feature based methodology utilizes key
the personnel staff ought to have an point features present on the face, called
appropriate system for approving and landmarks, of the face, for example,
maintaining the attendance record eyes, nose, mouth, edges or some other
consistently. By and large, there are two unique attributes. In this way, out of the
kinds of student attendance framework, picture that has been extricated
i.e. Manual Attendance System (MAS) beforehand, just some part is covered
and Automated Attendance System during the calculation process. Then
(AAS). Practically in MAS, the staff again, the brightness-based methodology
may experience difficulty in both consolidates and computes all parts of
approving and keeping up every the given picture. It is also called
student's record in a classroom all the holistic-based or image-based
time. In aclassroom with a high teacher- methodology. Since the overall picture
to-student ratio, it turns into an must be considered, the brightnessbased
extremely dreary and tedious process to methodology takes longer handling time
mark the attendance physically and and is likewise more complicated. There
cumulative attendance of each student. are different advances that are done
Consequently, we can execute a viable during the process of this face
framework which will mark the recognition framework, yet the essential
attendance of students automatically via steps of these are face detection and face
face recognition. AAS may decrease the recognition. Firstly, to mark the
managerial work of its staff. Especially, attendance, the images of students' faces
for an attendance system which will be required. This image can be
embraces Human Face Recognition captured from the camera, which will be
(HFR), it normally includes the students' installed in the classroom at a position
facial images captured at the time he/she from where the entire classroom is
is entering the classroom, or when visible. This image will be considered as
everyone is seated in the classroom to an input to the system. For efficient face
mark the attendance Generally, there are identification, the picture should be
two known methodologies to deal with upgraded by utilizing some image
HFR, one is the feature-based processing methods like grayscale
methodology and the other is the conversion and histogram equalization.
brightness-based methodology. The After image quality upgrade, the image

24
 ISSN 2347–3657
Volume 13, Issue 1, Jan 2025

will be passed to perform face detection. cumulative attendance of each student.

The face identification process is trailed Consequently, we can execute a viable
by face recognition process. There are framework which will mark the
different strategies accessible for face attendance of students automatically via
recognition like Eigen face, PCA and face recognition. AAS may decrease the
LDA hybrid algorithm. In the Eigen face, managerial work of its staff.
when faces are identified, they are
trimmed from the picture. With the III.PROPOSED SYSTEM
assistance of the element extractor, The proposed face recognition-based
different face highlights are extracted. attendance system leverages K-Nearest
Utilizing these faces as Eigen features, Neighbors (KNN) for real-time,
the student is recognized and by automated attendance management,
coordinating with the face database, specifically designed for educational
their attendance is marked. Developing institutions. This system begins with a
the face database is required with the data collection phase, where facial
end goal of comparison. images of students are captured and
stored in a structured database. During
II.EXISTING SYSTEM this stage, images undergo preprocessing
To verify the student attendance record, techniques such as grayscale conversion
the personnel staff ought to have an and histogram equalization to enhance
appropriate system for approving and facial features and optimize recognition
maintaining the attendance record accuracy. For real-time attendance, a
consistently. By and large, there are two live camera feed captures images as
kinds of student attendance framework, students enter the classroom. Using
i.e. Manual Attendance System (MAS) OpenCV’s Haar Cascade classifier, the
and Automated Attendance System system identifies faces within each
(AAS). Practically in MAS, the staff frame, cropping and resizing each
may experience difficulty in both detected face to a standard format. These
approving and keeping up every facial images are then converted into
student's record in a classroom all the feature vectors, which serve as input for
time. In aclassroom with a high teacher- the KNN classifier. By comparing these
to-student ratio, it turns into an vectors to existing records in the
extremely dreary and tedious process to database, the KNN algorithm determines
mark the attendance physically and the closest match based on Euclidean

25
 ISSN 2347–3657
Volume 13, Issue 1, Jan 2025

distance, effectively identifying each of 205 attempts, achieving a recognition

student in real-time. When a match is rate of approximately 85%.
confirmed, the system records the
2. P. Wagh, S. Patil, J. Chaudhari, and R.
student’s attendance automatically in an
Thakare focus on automating the
Excel file, updating the attendance
attendance system through face
database without requiring manual input.
recognition by using the Eigenface
database and PCA (Principal
IV.LITERATURE SURVEY
Component Analysis) algorithm in a
1.The paper by S. Lukas, A. R. Mitra, R.
Matlab GUI environment. Traditional
I. Desanti, and D. Krisnadi discusses the
attendance systems face several issues,
implementation of a student attendance
such as fake attendance, time
system in the classroom using face
consumption, and manipulation, while
recognition techniques. In recent years,
also lacking secure information
biometric authentication methods such
management. However, face recognition
as fingerprint recognition have become
technologies face challenges like image
popular in various applications like
quality, size, angle, and varying light
video surveillance, human-computer
intensity. To overcome these limitations,
interaction, door access control systems,
the authors use techniques like
and network security. While fingerprint
Illumination Invariant, Histogram
recognition is considered one of the
Equalization, and PCA. The proposed
most accurate and fast methods due to
system automatically updates attendance
its uniqueness and permanence, face
after comparing the detected face with
recognition is preferred for capturing
the original Eigen database integrated
student attendance in a classroom setting.
with an Excel sheet and Matlab GUI.
In addition to attendance, face
recognition can also provide insights
3. The paper by Viola, M. J. Jones, and
into student behavior, such as their
Pau presents a face detection framework
readiness or interest in the lecture. The
designed for high-speed processing
proposed method uses multiple facial
while maintaining high detection rates.
images to classify facial objects.
The authors introduce a novel image
Experiments were conducted with 19
representation called the "Integral
students in a classroom setting, resulting
Image," which allows for rapid
in 174 successful face recognitions out
computation of features used by the

26
 ISSN 2347–3657
Volume 13, Issue 1, Jan 2025

detector. Additionally, the framework added to the student database, where

employs a simple and efficient classifier general information such as name and
built using the AdaBoost learning enrollment number is stored.
algorithm to select a small number of Additionally, images of the student's
critical visual features from a large set. face captured by the camera are stored in
The paper also introduces a method for the database. These images are critical
combining classifiers in a "cascade," for facial recognition and allow the
enabling faster background elimination system to identify and verify students
while focusing computational resources attending lectures. By using the stored
on promising face-like regions. The facial images, the system can recognize
system's performance is evaluated in the students as they enter the classroom and
domain of face detection, yielding perform attendance tracking.
comparable results to previous systems.
Face detection is a crucial first step in
When implemented on a conventional
this face recognition-based attendance
desktop, the system processes face
system. It is responsible for accurately
detection at 15 frames per second.
identifying faces in a live video feed.
The system uses OpenCV’s Haar
Cascade Classifier to perform real-time
face detection. The classifier employs
pre-trained data to recognize key facial
features, such as eyes, nose, and mouth.
Once the camera captures an image, the
system detects the face and highlights it
Fig1: System Architecture
by drawing a bounding box around it.
This process makes it easy to monitor
V.METHODOLOGY
the success of each detection. The
To develop the smart attendance
detected face is then cropped and resized,
management system, several steps need
preparing it for feature extraction and
to be followed to ensure successful
recognition by the KNN classifier. Haar
implementation. These steps are as
Cascade offers a balance of speed and
follows: Enrollment, Face Detection,
accuracy, making it ideal for real-time
Face Recognition, Confirmation by the
attendance tracking. Face recognition is
class camera, and Attendance Marking.
implemented using Principal Component
In the enrollment step, the student is

27
 ISSN 2347–3657
Volume 13, Issue 1, Jan 2025

Analysis (PCA), a technique for the attendance database, finalizing the

reducing the number of variables process for that particular lecture.
involved in facial recognition. In PCA, OpenCV is a powerful, cross-platform
each image in the training dataset is library used to develop real-time
represented as an eigenvector, referred computer vision applications. It focuses
to as an "eigenface." This methodology on image processing, video capture, and
transforms faces into a smaller set of analysis, offering various features such
features—eigenfaces—which are the as face detection and object detection.
principal components of the original OpenCV simplifies the development of
training images. Recognition is applications related to real-time visual
performed by projecting a new image processing. Computer vision is a field
into the eigenface subspace and that deals with the interpretation and
comparing its position with the positions understanding of 3D scenes from their
of known individuals. The advantages of 2D images, with a focus on modeling
using PCA for facial recognition include and replicating human vision through
its ease of use, speed, and stability computer software and hardware. It
despite variations in human faces. When significantly overlaps with areas such as
a student approaches the classroom, image processing, which focuses on
their face is checked against the image manipulation; pattern recognition,
database. If their face matches an entry, which involves classifying patterns; and
they are granted access; if not, they are photogrammetry, which is concerned
prompted to enroll in the system. After with extracting accurate measurements
face recognition, the system employs a from images. The distinction between
confirmation step using a second camera computer vision and image processing
installed inside the classroom. This lies in their objectives. Image processing
camera is strategically placed to ensure involves transforming images, where
all students are visible, helping to both the input and output are images. In
confirm their presence in the class and contrast, computer vision goes a step
preventing proxy attendance. This step further by creating meaningful
ensures the accuracy of the attendance descriptions or interpretations of 3D
records. At the end of the lecture, the scenes from their 2D images. The output
camera inside the classroom is used to of computer vision is a description or
generate a list of students present. This understanding of the physical structure
list is then used to mark attendance in represented in the scene.

28
 ISSN 2347–3657
Volume 13, Issue 1, Jan 2025

faces at one instance. Also, there is no

requirement of any special hardware for
its implementation. A camera, a PC and
database servers are sufficient for
constructing the smart attendance
system.
Fig2: Mark Attendence

VII.REFERENCES
1. S. Lukas, A. R. Mitra, R. I. Desanti
and D. Krisnadi, "Student Attendance
System in Classroom Using Face
Recognition Technique," in ICTC 2016,
Karawaci, 2016.
2. P. Wagh, S. Patil, J. Chaudhari and R.
Fig3: View Attendence
Thakare, "Attendance System based on
Face Recognition using Eigen face and
VI. CONCLUSION
PCA Algorithms," in 2015 International
The proposed automated attendance
Conference on Green Computing and
system using face recognition is a great
Internet of Things (ICGCloT), 2015.
model for marking the attendance of
3. N. M. Ara, N. S. Simul and M. S.
students in a classroom. This system
Islam, "Convolutional Neural Network
also assists in overcoming the chances
Approach for Vision Based Student
of proxies and fake attendance. In the
Recognition System," in 2017 20th
modern world, a large number of
International Conference of Computer
systems using biometrics are available.
and Information Technology (ICCIT),
However, the facial recognition turns
22-24 December, 2017, Sylhet, 2017.
out to be a viable option because of its
4. N. Khan and Balcoh, "Algorithm for
high accuracy along with minimum
efficient attendance management: Face
human intervention. This system is
recognition based approach," in JCSI
aimed at providing a significant level of
International Journal of Computer
security. Hence, a highly pro-efficient
Science Issues 9.4, 2012.
attendance system for classroom
5. KAWAGUCHI and Yohei, "Face
attendance needs to be developed which
Recognition-based Lecture Attendance
can perform recognition on multiple

29
 ISSN 2347–3657
Volume 13, Issue 1, Jan 2025

System.," in The 3rd AEARU Workshop OpenCV," in International Conference

on Network Education. 2005., 2005. on Electronics, Communication and
6. MuthuKalyani.K, "Smart Application Aerospace Technology, ICECA 2017,
For AMS using Face Recognition," in 2017.
CSEIJ 2013, 2013. 9. Viola, M. J. Jones and Paul, "Robust
7. M. Arsenovic, S. Skadojevic and A. real-time face detection.," in
Anderla, "FaceTime- Deep Learning International journal of computer vision
Based Face Recognition Attendamce 57.2 (2004), 2004.
system.," in IEEE 15th International 10. A. Jha, ""Class Room Attendance
Symposium on Intelligent Systems and System Using Facial Recognition
Informatics, Serbia, 2017. System."," in International journal of
8. K. Goyal, K. Agarwal and R. Kumar, Mathematical science technology and
"Face Detection and tracking using management 2(3). 2007, 2007.

30