SYNOPSIS ON

SMART ATTENDANCE MANAGEMENT SYSTEM

GURU TEGH BAHADUR INSTITUTE OF TECHNOLOGY

SUBMITTED BY –

PETER BAGCHI(01013207721)
ANKUR CHAUDHARY(00913207721)
FAIZAN AHEMAD(01213207721)

MENTOR-
Abstract 1.2. FACE RECOGNITION SYSTEM

Managing the attendance can be a great burden if teachers take Face Recognition is a process of marking attendance of
attendance manually by calling out each and every student each student by recognizing the face of the student to
name. To resolve this problem, smart attendance management ensure that his/her presence in the classroom. During
system is being used. Each person has an Unique face. In this the lectures it can also be used for the lecturer to find
project face of an individual is used for the purpose of whether the student is sleeping or awake during the
automatically marking the attendance. The smart attendance lectures.
management system is executed generally with the help of
biometrics. One of the biometric method called Face recognition A web cameras or High-Definition camera is used to
is used to improve this system. Facial recognition is being used capture the image of the student by using the installed
vastly in many applications, like System of CCTV footage and cameras in each classroom. Thus from the captured
video monitoring, an interaction between humans and computers. image, the attendance is marked for the student in the
Traditionally, attendance is taken by calling the name or roll database, based on face recognition of each student the
number of the student andthen marked attendance is recorded. images are further segregated by recognizing the face
Time consumption is themain concern for this purpose. Suppose region based on the captured image to ensure their
the duration for one subject is around 60 mins, it takes about 5
to 10 minutes to record attendance. This leads in excess time presence in the classroom. By using some machine
consumption. To prevent such losses, an automatic process learning algorithms, the attendance marking in the
attendance marking is used in this project which is based on database is done based on the face segregation of the
image processing. Face detection and face recognition is used in student. The Face Recognition technology is considered
thisproject because without any influence of either the student orthe
lecturer or any other external person, the attendance of
Key Words: Face Recognition, Machine Learning, Open CV,
Support Vector Machine.
each student can be easily marked based on the face
image of the student that can be captured from a various
INTRODUCTION locations and which can be further usedfor analysis. As a
An Attendance plays a vital role in the educational organization result, no student can put the proxy attendance and
for both the teacher and student. As it plays a prime important imitate as another student. As now a days everything is
role to keep the record of the attendance. While earlier, as we in biometric form such as the physical fingerprint of the
think about the traditional process of taking the attendance in student is required to mark the attendance based on
class room by Calling name or roll number of each student for considering fingerprintrecognition system or, iris of an
the attendance, the problem arises by following this traditional individual student is supposed to be scanned through a
process is more time consumption and also it requires more biometric device to mark attendance based on
energy. As the students have to wait in queue for giving their considering iris recognitionsystem. If we considered the
attendance, which is very much time consuming. In order to other recognition technologies which requires the
overcome with this problem an automatic attendance physical interferenceof each and every student whereas
management system is designed, that can solve all the above the face recognition technique does not require any
related problems.
physical interference of the student. Therefore, on-going
Currently there are so many institutions and organizations who with a face recognitionsystem, the face of the student can
are using automatic attendance managing system. One of the be easily captured and obtained through the camera
most commonly used system in these institutions and installed in the classroom and the attendance of each
organizations is biometric technique form. As this technique is student can be easily maintained in a database. The face
automatic and a step forward compared to traditional method, recognition system contains two major things that need
which has been failed to meet the time constraint. to bemaintained as follows,
This project helps by introducing an involuntary attendance i. Enhanced Security - because the face image of
every student should not be misused by an
marking system, devoid of any kind of interference with the
external authority.
normal teaching procedure. The system can also be used by
ii. Faster Processing-helps in maintaining
implementing the cameras during exam sessions or in other
attendance for all the students of the institution,
teaching activities where attendance is highly mandatory. By thus faster processing helps for easier retrieval of
using this system which helps to eliminate the classical student attendance to send email or message to the parent.
identification such as calling name of the student, or checking
respective identification cards of the student, which can not
only interfere with the ongoing teaching process, but also can
be stressful for students during examination sessions.
 1. PROPOSED SYSTEM 1.2. Face Recognition
The proposed system task is to capture the face of each The face image detected in each frame is taken and
individual student and to store it in the database for compared with the directory where the pre-trained
students attendance. The student face needs to be face image of the student has resided. Again the same
captured in such a way that all the students face features process is repeated for all the frames. To perform this
needs to be detected and even the posture andthe seating comparison the process of Super Vector Machine
of the student needs to be recognized.There is no need Learning algorithm is employed.
for the teacher to take the attendance manually in the
class because the system records a video and through 1.3. Attendance Marking
If face image presenting the frame1matches the pre-
further processing steps the face isbeing recognized and
trained image the n the attendance is obtained of the
finally the attendance database isupdated.
particular student for the concerned lecture hour. If the
OpenCV: OpenCV (Open Source Computer Vision student named as frame1 and that is not available in the
Library) is a library of programming functions mainly trained dataset then student's face saved as
aimed at real-time computer vision. OpenCV is image1doesn’t belong to the particular class, which
developed by Intel and later supported by WillowGarage means that the particular student may belong to the
different class.
and then Itseez. The library is free for use underthe open-
source BSD license. Python is dynamically typed and 2. FUTURE WORK
garbage is collected. It supports multiple programming i. Future work of this project can be sending alert
paradigms, including structured, object- oriented and SMS to student with reference of the attendance.
functional programming. Python is often described as a GSM module can be used for this concept. Alert
"batteries included" language because of its SMS can be sent to the parent of the student.
comprehensive standard library. The proposed face ii. The accuracy level of this project has reached up
recognition is categorized into four major steps namely, to 80% for imperfect and opaque images. Further
it can be improved for obtaining higher accuracy
i. Image Acquisition levels.
ii. Face Detection iii. Two or more IP cameras can be further used for
iii. Face Recognition working and processing of each image can be
iv. Attendance Marking done separately. The merged results of these can
1.1. Image Acquisition help to obtain better results and accuracy in
The first step involved in the second phase is image denser classrooms.
acquisition, that is obtaining the face image of the REFERENCES
[1] Bhumika G. Bhatt, Zankhana H. Shah “Face Feature
students which is present in the classroom. It can be
Extraction Techniques: A Survey”, National Conference
which will be obtained during the lecture hour, the frames
of each sequence are extracted fromthe video and on Recent Trends in Engineering & Technology, 13-14
numbered for further processing obtained through the May 2011.
high Definition Video Camera that installed in each [2] Kar, Nirmalya, et al. "Study of implementing
classroom.
automated attendance system using face recognition
technique." International Journal of computer and
communication engineering 1.2 (2012): 100.
1.1. Face Detection
From the frames which are extracted, each face of the
image needs to be segregated (divided).For this
segregation purpose, we go with the face region bounding
box methodology process usually it is calledmarking the
region of interest using HAAR cascade classifiers that is
available in the OpenCV and face- recognition library
respectively. After dividing the frame, the first frame is
taken and the face image is detected and then it is
marked. Then next the second frame is taken and again the
face image is detected andmarked. Again the same process
is repeated for all theframes.