IJCSN - International Journal of Computer Science and Network, Volume 8, Issue 2, April 2019

ISSN (Online) : 2277-5420

www.IJCSN.org
Impact Factor: 1.5

Automatic Attendance Management System

Using Face Recognition
1
Swapna Munigala; 2 Samiha Mirza; 3 Zeba Naseem Fathima; 4 ZubairaMaheen
1
CSE Department, Osmania University- Stanley college of Engineering and Technology for Women,
Asst Prof CSE Dept, Hyderabad, Telangana 500036, India
2
CSE Department, Osmania University- Stanley college of Engineering and Technology for Women,
B.E IV Year, Hyderabad, Telangana 500036, India
3
CSE Department, Osmania University- Stanley college of Engineering and Technology for Women,
B.E IV Year, Hyderabad, Telangana 500036, India
4
CSE Department, Osmania University- Stanley college of Engineering and Technology for Women,
B.E IV Year, Hyderabad, Telangana 500036, India

Abstract - Student performance is analyzed by their attendance. Nowadays manual attendance is considered as a time consuming job
and one might lose the copy of attendance. So instead of taking it manually we refer to take in biometric, on the whole we want the
total work to be an automated process. For it, to be automated the main goal is to recognize and verify a person’s face from the
screenshot captured, because a major part comes with the individual faces and then the information is checked in file along with marks
the attendance for student. Traditional face recognition systems take up a method to identify a face from the given input, but in case the
output is not accurate and sometimes is hazy. In this paper our aim is to deviate from such traditional systems and introduce a new
approach to identify a student using Face Recognition System. There are many techniques which can be executed with it such as Multi
Linear subspace learning using Tensor representation, hidden Markov model etc.

Keywords - Automatic face Detection, Face Identification, Face recognition, Feature Extraction, Face Matching.

1. Introduction There are loads of self-determining methods on hand for

this point i.e. biometric attendance. All these methods
rtificial Intelligence: It is also machine deplete time because students have to make a queue to
touch their thumb in the scanning device. Main aspire of

A intelligence which is signified by machines, in

computer science it is research defined as
intelligent agents which perceives the
environment by taking actions and achieving
the goals. This paper is under the domain of Artificial
this project is to make whole work automated. It describes
the well-organized algorithm that without human
intervention marks the attendance without human
interposing. These attendances is recorded by passing the
Intelligence, it is conservatively associated with input externally using a camera attached in front of the
security sector. Making us to evolve with an AI based lecture room that is continuously capturing images of
application that uniquely identifies a person by students, detects the faces in images and compare the
examining patterns based on facial textures. It is an detected faces with the database, by following the
rational task where our goals is to computerize the techniques and marking the attendance. The paper estimate
whole work. Everyday actions are increasingly urging the correlated work in the field of attendance system then
to handle by machine, as an alternative of pencil and describes the system structural design, software algorithm
paper or face to face development in electronic and results.
transactions outcome in great demand for swift and
exact user identification and authentication. For new Biometric is a unique quantifiable characteristic of a [2]
things to add [1] we in this project want a system human being that can be used automatically to recognize an
which takes attendance automatically by capturing individual or verify individuals in identity. Biometrics can
image and featuring attendance by searching in determine both physiological (based on from a direct
database. Students attendance in the classroom is vital measurement from a part of human body) and behavioral (is
duty and if taken hand operated wastes a lot of instant. based on measurements and data derived from an action)

172

Copyright (c) 2019 International Journal of Computer Science and Network. All Rights Reserved.
IJCSN - International Journal of Computer Science and Network, Volume 8, Issue 2, April 2019
ISSN (Online) : 2277-5420
www.IJCSN.org
Impact Factor: 1.5

characteristics. Face recognition has the capacity to become an

invaluable part of many identification systems used for
2. Literary Survey evaluating the performance of those people functioning
within the organization. Biometric technologies are being
As traditionally the areas under this paper majorly Face appertained in many fields which is not yet delivered, its
Recognition has been an active research area over the assure of guaranteeing automatic human recognition. Face
last decades. There are several sub methods like [11] recognition is a technique of biometric recognition. It is
Image Processing, Machine learning approach, Pattern considered to be one of [4] the most affluent applications
Recognition etc. In the process of recognition it of image analysis and processing which is the main reason
comprises, to identify the images individually and then behind the great attention it has been given in the past
categorize the new coming test cases. several years.

Fig 1. Flowchart for Attendance System

two main process: organizing before detection where face
As the face is the main part of an individual to detection and alignment takes place (localization and
recognize and this becomes the input for the project to normalization) afterwards recognition transpire through
go ahead. One way is comparing selected features feature extraction and matching steps. There are various
from the image and identifying from facial database. techniques like Principal component Analysis (PCA),
For attendance this is secured way of marking Linear Discriminate Analysis(LDA) , Elastic Bunch Graph
compared t other biometric systems. Some algorithms Matching using Fisher face algorithm, Neuronal Motivated
do their work by extracting special features like by Dynamic Link Matching, this is evident by numerous face
differing sizes, color, iris etc from the given input. [6] recognition researches. This research is to propose and
Preferably the facial recognition process can split into enhance pre processing algorithms and follow in this

173

Copyright (c) 2019 International Journal of Computer Science and Network. All Rights Reserved.
IJCSN - International Journal of Computer Science and Network, Volume 8, Issue 2, April 2019
ISSN (Online) : 2277-5420
www.IJCSN.org
Impact Factor: 1.5

project. given which derives building features for informative,

facilitating, learning and also for better interpretations. It
3. Automatic Face Detection connects to dimensionality reduction. This comes when the
data is large to be processed it is suspected to get
It is a biometric allied computer technology being used transformed in reduced sets of features. The opted feature
in diverse applications which differentiate human faces for this contains the pertinent information from the input
in digital images. It also direct to the physiological data, so that the desired task can be performed by using this
process by which humans locate and deal to faces in a reduced representation instead of complete initial data.
[16] visual scene. It is used as human computer Basically in this we have things to concentrate; as our
interface and database management. A face Detector features are easy to calculate but we need to take off the
has to tell whether an image of random size contains a white spaces, then we generate an integral image which best
human face and if so where it is. Face detection can be suits for feature extraction making the process fast. So
performed based on several prompts: skin color, facial features are extracted from sub windows of the given
/head shape, facial appearance or a combination of image. The future of this could lead us to numerous data.
these parameters . Most face detection algorithms are But the face detection step, patches are extracted from
appearance based without using other cues. An input images. [13] Performance of patches have some downsides
image is scanned at all possible locations by a sub like using this patches directly for face recognition like
window. Face detection is constituted as classifying the firstly, each patch generally contains over 1000 pixels,
pattern in the sub aperture either as a face or a non which are substantial to build a vigorous recognition
face. system. Secondly, face patches may be taken from different
alignments, unlike face expressions, illuminations, but may
4. Face Identification suffer from ceasing and jumble. To solve these drawbacks,
feature extractions are performed so to information packing,
It is also a technology which is efficient of identifying dimension reduction, salience extraction, and cleaning.
or verifying a person from the digital image. Henceforward, the patch is modified into a vector of fixed
Importantly it is an Biometric Artificial Intelligence dimension or a set of fiduciary points with their
based application that uniquely identifies a person by corresponding locations. As the given pattern by applying
analyzing pattern based on person’s textual features, techniques conventionally we get to extract the features and
size, extracting landmarks etc. There are numerous then decision procedure with intense classification. In some
methods in which facial recognition system works but papers, feature extraction is either included in face
usually people work using a selected facial feature and detection or face recognition.
comparing them from the input given. It is like outer
perception by which human brain understands and 6. Face Matching
interprets the face. Popular recognition algorithms
include PCA, LDA and Multi linear subspace learning The features extracted are then contrasted to those in data
using tensor representation. It is however be noted that and decisions are made seeing those. Then this matching
the existing face identification techniques are not 100% technique has various applications in:
capable yet. Typical efficiencies range between 40 to 1. Video Gaming
60 percent. By this no manual contact on behalf of the 2. Face ID
user is required. It is only biometric that allow us to 3. User Authentications
perform passive identification in a one to many 4. Registrations
environments. 5. Security
6. Law and video Surveillance
7. Others
While matching this makes the job easy and then stores in
database. Comparison [17] is then over with the progress of
data marking of attendance.

Fig 2. Procedural Steps 7.Problem Definition

5. Feature Extraction
Taking attendance is a tedious process and takes lot of
It involves reducing the amount of resources. Feature exertion and time, especially for enormous number of
extraction starts from an initial set measured from data students. It is also tricky when an exam is held and causes a

174

Copyright (c) 2019 International Journal of Computer Science and Network. All Rights Reserved.
IJCSN - International Journal of Computer Science and Network, Volume 8, Issue 2, April 2019
ISSN (Online) : 2277-5420
www.IJCSN.org
Impact Factor: 1.5

lot of commotion. As it is a manually done the 10. Sample Outputs

attendance sheet is subjected to damage and loss while
being passed on between different students When the
students enrolled for a certain course is massive, the
lecturers call them by names arbitrarily which is not
reasonable for student assessment process either. This
procedure could be easy and effectual with a minute
number of students but on the other hand dealing with
the records of a huge number of students often leads to
human error.

8. Proposed System
We propose solution to all the mentioned problems by
proffering an automated attendance system for all the
students who are present for a monologue, section,
laboratory or practical’s at its specific time
consequently saving time, attempt and disturbance.
Another thing concerns exams, is when the lecturer or Fig3. GUI Screen
the advisor accidentally looses an exam paper or the
student lies about attending the exam, there will be a
proof of the students attendance for the exam at that
time, thus protecting both lecturer’s and students’
legality. In addition, an automated evaluation would
provide more precise and dependable results avoiding
human error. The main purpose of the system is to
make the total work an automated process which is
active, reliable, eliminating disturbance, time loss in
traditional attendance systems and giving accurate
students performances. The system consists of a
camera that captures the picture of the classroom at the
incept of the session. Next the processing module
image is improved to smooth the progress of the
matching process. After this face detection and
recognition is performed. At this seam both the images
are juxtaposed and the students who exist in both the
images are marked present in the database.
Fig4. Marking attendance for Face 1

9. Implementation
We carried out techniques like face detection using
Haar Feature-based Cascade Classifiers for face
detection. In detection and for feature extraction, Local
binary patterns (LBP) which is a kind of illustration
descriptor for categorization in computer vision. Each
front (face) has a unique histogram. This test image
histogram is compared with train image histogram for
similitude. LBP is the particular type of the Texture
field model propounded in 1990 and eventually charted
and modernized. This is combined with the Histogram
of oriented gradients (HOG) descriptor, which
improves the detection staging noticeably on some
datasets. Fig5. Marking attendance for Face 2

175

Copyright (c) 2019 International Journal of Computer Science and Network. All Rights Reserved.
IJCSN - International Journal of Computer Science and Network, Volume 8, Issue 2, April 2019
ISSN (Online) : 2277-5420
www.IJCSN.org
Impact Factor: 1.5

[2] Venkata Kalyan Polamarasetty and Muralidhar Reddy

Reddem, “Attendance System based on Face Recognition”
in proc.IRJET,Apr2018.
[3] Naveed Khan Baloch and M.Haroon Yousaf, “Algorithm
for Efficient Attendance Management: Face Recognition
based apporach” in IJCSI, July 2012.
[4] Abhishek Jha, “Class Room Attendance System Using
Facial Recognition System” in proc. Journal of
Mathematics Science,Technology and Management.
[5] Ashish Choudhary and Abhishek Tripathi, “Automatic
Attendance System Using Face Recognition” in proc.
IJMTER, April 2016.
[6] Mr.C.S.Patil , Mr.R.R.Karhe and Mr.M.D.Jain, “Student
Attendance System and Authentication using Face
Recognition” in proc. IJERT ,July-2014.
[7] Ajinkya Patil and Mrudang Shukla, “Implementation of
Classroom Attendance System based on Face Recognition
in Class” in proc. IJAET,July-2014 .
[8] Mathana Gopala Krishnan and Balaji, “Implementation of
Automated Attendance System Using Face Recognition”
Fig6. Marking attendance for Face 3 in proc. Journal of Scientific & Engineering
Research,March-2015.
[9] Jomon Joseph and K.P.Zacharia “Automatic Attendance
Management System Using Face Recognition” in proc.
IJSR,November2013.
[10] Viola P. & Jones J…Robust Real-Time Face Detection
International Journal of Computer Vision 57(2), 2004.
[11] Bhumika G. Bhatt, Zankhana H. shah “Face Feature
Extraction Techniques: A Survey”, National conference on
Recent Trends in Engineering & Technology, May 2011.
[12] M.A. Turk and A.P. Pentland, “Face Recognition Using
Eigenfaces” in proc. IEEE Conference on Computer
Vision and Pattern Recognition, 1991.
[13] J.A.C. Weaver, “Biometric authentication”,
Computer,2006.
[14] W. Zhoa, R. Chellappa, P.J. Phillips, and A. Rosenfield,
“Face recognition: A literature survey”, ACM Computing
Fig7. Final Attendance Sheet Surveys, 2003, vol.35.
[15] Y. Cui, J.S. Jin, S. Luo, M. Park, and S.S. L.. Au,
11. Future Scopes and Other Remarks “Automated Pattern Recognition System”, in proc. 5 th
International Conference on Computer Vision and
The system we have developed is successfully able to Graphical Image, vol.59, May 1992.
[16] Ms. Varsha Gupta, Mr. Dipesh Sharma, “A Study of
accomplish the task of marking the attendance in the
Various Face Detection Methods”, in proc IJARCCE
classroom automatically and output obtained is updated vol.3, May-2014.
as programmed. On another aspect where we could try [17] Anil K. Jain, Arun Ross and Salil Prabhakar, “An
is creating an online database of the attendance and its Introduction to Biometric Recognition”, IEEE
self-regulating updates, keeping in mind about the Transactions on vol 14, Issue, Jan. 2004.
growth of internet of things. [18] Onur Sanli, Bahar Ilgen “Face Detection and Recognition
for Automatic Attendance System” Springer-Verleg, Nov
2018.
References [19] Viola Jones algorithm, PCA algorithm and related content
from Wikipedia [en.wikipedia.org].
[1] NaveshSallawar and Shubham yende,“Automatic
attendance system by using face recognition” in proc.
IRJET, Apr-2017.

176

Copyright (c) 2019 International Journal of Computer Science and Network. All Rights Reserved.