International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056

Volume: 05 Issue: 04 | Apr-2018 www.irjet.net p-ISSN: 2395-0072

Attendance System based on Face Recognition

Venkata Kalyan Polamarasetty1, Muralidhar Reddy Reddem2, Dheeraj Ravi3,
Mahith Sai Madala4

1,2,3,4Student, Information Technology, Velagapudi RamaKrishna Siddhartha Engineering College, India

---------------------------------------------------------------------***---------------------------------------------------------------------
ABSTRACT: Daily attendance marking is a common and training data sets. Instances taking camera capture now
important activity in schools and colleges for checking the check that image to database Images. Face recognition of
performance of students. Manual Attendance maintaining is different peoples based on the related images of that person
difficult process, especially for large group of students. Some image we need take images for before face recognition. In
automated systems developed to overcome these difficulties, case if the image is not in data base then we store that image
have drawbacks like cost, fake attendance, accuracy, as new person in database. Next time same image of that
intrusiveness. To overcome these drawbacks, there is need of new image person appear in image and recognition the face
smart and automated attendance system. Traditional face or else taking as new image and storing in database process
recognition systems employ methods to identify a face from is repeating. In this paper we selecting of the face
the given input but the results are not usually accurate and recognition and detection giving result using MATLAB. This
precise as desired. The system described in this we aims to requires a high end specifications of a system in order get
deviate from such traditional systems and introduce a new the better results. It won’t run on all the small specification
approach to identify a student using a face recognition system, systems. So, this can run only small database and compare
the generation of a facial Model. This describes the working of them with the face required.
the face recognition system that will be deployed as an
Automated Attendance System in a classroom environment. 2. LITERATURE SURVEY

Keywords: Image processing, Expressions, HAAR features, In [1] the author proposed that different types of face
HOG features, Face Detection Method, Face Recognition detection for detecting faces in different pose .Detecting face
Method, Features Extraction, Features Matching, Query Image, in different pattern based on techniques. Basic pattern for
Bit- Byte Conversion Methods. detecting face is nose, eyes, hair, ears and some time it based
on tone of skin. Face detection is detecting face based on
1. INTRODUCTION location of face and presences of face in images .Different
types of detecting the face techniques they are Ada-Boost
Let us take an example of application to the theory we are Algorithm for Face Detection, Viola Jones Face Detection
proposing here. Taking attendance in the schools and Algorithm, SMQT Features and SNOW Classifier Method,
colleges is being a waste of time and effort for both the Local Binary Pattern (LBP). Each have advantages and
students and lectures as well. Now a days biometric is more disadvantages discussed in that paper.
usage they are finger print recognition facial recognition iris
scanning recognition voice recognition signature recognition Xiang-Yu Li [2] the author proposed that recognition face
etc. One of that biometric category is face detection and using hog features and pca algorithms. By applying
recognition. Based on the image we take security safety, 0recognition algorithm to cropped faces images from that we
attendances and some time it useful for decision also. Mostly get similarity b/w taken image and database image. In this
this facial detection and recognition is decrease the manual paper PAC algorithm used for face detection and recognition.
work for human. Image capturing from camera or cc camera Arun Katara[3] the author shows that face recognition of
sometime this is also a streaming video from camera. Form facial of different person or student .from recognition
that offline or online data, we capture the image after that attendances is upload to database using face detection and
applying the face detection techniques. Face detection is recognition of student or workers. From this manual work is
detecting the face location and presence of face in images. In decrease by human and automatically attendance system
this face detection we mostly see the nose, hair, ears, mouth, based on faces process done.
eyes and also different pose of faces in images. So many Face
detection techniques, few of them is Viola Jones Face In [12] authors have consider a system based on real time
Detection Algorithm, (LBP), and Ada-Boost for Face face recognition which is fast & which needs improvisation
Detection, SMQT Features and SNOW Classifier Method. of images in various lighting environments.
After applying face detection techniques we detected the
faces or objects in image and crop that image apply Face
recognition technique. So many way to recognition the faces
by applying Hog features, Haar features, Machine learning,
deep leaning, classification techniques some other tech also
used for recognition of the faces. Recognition of face we need

© 2018, IRJET | Impact Factor value: 6.171 | ISO 9001:2008 Certified Journal | Page 4606
 International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 05 Issue: 04 | Apr-2018 www.irjet.net p-ISSN: 2395-0072

3. Work Flow Now we need to split the data of each and every person into
testing and training data. Let us take it in the ratio of 0.2:0.8
from the database. Now we will be extracting the HOG
features of all the training individuals and store them in the
form of bits and bytes. We need to fetch the cropped and
gray scaled images. Now the training datasets are extracted
with the HOG features and are stored with a count.

Fig-3.1 Architectural Flow Chart

3.1 Training data

For the database, we need to capture the image from the

webcam or the external camera. To do so, in Matlab, we need
to install the drivers from the math works website based on
the type of camera we are using. Next, we need at least 500
to 1000 captures of each person for getting higher Fig-3.4.1 Hog Feature of the Input image
percentage of accuracy and meet the purpose we are doing
in this project. We can store the data in the form of separate All the HOG features extracted are stored in the form of array
folders distinguishing each person from others. index. The images in the database along with their labels are
sent into features array to identify them separately (like
3.2 Image Capture indexing). The HOG features are sent along with the person
Label to classify them and store them separately. The data is
We need some HD camera in order to get results. We can now classified using the predefined method of fitcecoc. This
capture the images from the video stream or by capturing is completely done on the training data. Now we need take a
each and every image from the webcam manually. Doing the fresh photo from webcam or any file and detect the faces,
frame capture from the stream of video will give us results in extract the HOG features and then compare that with the
less time but we won’t be able to capture the face properly in data classified. We need a predict method to compare an
case we lose light or something and if the face is not classified data with the data we need. Finally it returns a
captured properly. label to which the given data matches or nearly matched. To
get better results we need a bigger training set.
3.3 Face detection
3.6 Front end
For face detecting, we can do it using the object cascading
class and we use the b-box method. The detection of the face In Matlab, we use an unique tool to design the user interface.
using the object cascading is bought from the most popular Here we can create buttons and plots in order to get the
facial recognition model Viola Jones. In here, there are image frame from the webcam. We can allot some callback
several objects are present. These are there in the form of functions in order to configure the camera from which we
small blocks containing them. They are taken through an want take the image, capturing the image and saving the
image and are moved through each and every block of the image. We can also use the edit boxes and static boxes in
image and are checked for overlapping through them. First order to take the data from the user and display some
we will convert the image from the red blue green to gray content to the end users.
scale image. The faces from the image captured is to be
collected. The captured faces are cropped into small images 4. Result Analysis
of resolution 112x92. It would be around 11 KB of size.
First in all we need to register the person into the database.
3.4 Face Recognition To do so, we need to give name and his/her registered
number to store.
The faces taken in the database are needed to loaded into
our workspace. We will load the galleryimages into that.

© 2018, IRJET | Impact Factor value: 6.171 | ISO 9001:2008 Certified Journal | Page 4607
 International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 05 Issue: 04 | Apr-2018 www.irjet.net p-ISSN: 2395-0072

Fig-4.1 Registration form

Now we to get the pictures of the persons from the webcam

or any other cams available like we used Logitech webcam Fig-4.3 Interface to take attendance
here. Select the cam from which we need to take the image
and start the camera. The camera is plotted in the axes and Now let us click on the “Take Attendnce” and the camera
we can capture and save the images in the folder created starts and takes the image to give the results checking from
automatically with the registered number we have entered the saved database
in fig 4.1.

Fig-4.2 User Interface to select, capture and save the

images in database

After this the data is stored in the database. Now let us

capture a picture from the webcam and see the results.

Fig-4.4 Input image

The input image Fig 4.4 now undergoes the predict process
and the name of each person are given as in Fig 4.5

© 2018, IRJET | Impact Factor value: 6.171 | ISO 9001:2008 Certified Journal | Page 4608
 International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 05 Issue: 04 | Apr-2018 www.irjet.net p-ISSN: 2395-0072

Fig-4.6 View Attendance

5. Conclusion

Capturing the images from camera or cc camera and

applying techniques face detection and recognition can
decrease the manual work from human and increase the
security safety, taking the decision from this recognition
result. Based on this face detection and recognition can used
in implement so many application like automatic
attendances system based on face recognition, worker
attendances, security, safety, police application like finding
thief in image that help to catching thief. In this system we
have implemented an attendance system for a lecture,
section or laboratory by which lecturer or teaching assistant
an record student’s attendance. It saves time and effort,
especially if it is a lecture with huge number of students. The
complete system is implemented in MATLAB. This
attendance system shows the use of facial recognition
techniques for the purpose of student attendance and for the
further process this record of student can be used in exam
related issues.

References
Fig-4.5 Output for the image in Fig 4.4
[1]A Study of Various Face Detection Methods , Ms.Varsha
Found images are given their names and not found ones
Gupta1 , Mr. Dipesh Sharma2,ijarcce vol.3
are as the image in the middle.
https://www.ijarcce.com/upload/2014/may/IJARCCE7G%2
0%20a%20varsha%20A%20Study%20of%20Various%20F
View Attendance ace.pdf
If we click on the View Attendance button in fig-4.3 we will [2]Face Recognition Based on HOG and Fast PCA Algorithm
logout as shown in fig-4.6. The attendance is viewed after Xiang-Yu Li(&) and Zhen-Xian Lin
capturing of the fig-4.. Initially all the members are marked
absent.

© 2018, IRJET | Impact Factor value: 6.171 | ISO 9001:2008 Certified Journal | Page 4609
 International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 05 Issue: 04 | Apr-2018 www.irjet.net p-ISSN: 2395-0072

[3]Attendance System Using Face Recognition and Class

Monitoring System, Arun Katara1, Mr. Sudesh2,
V.Kolhe3http://www.ijritcc.org/download/browse/Volume_
5_Issues/February_17_Volume_5_Issue_2/1489565866_ 15-
03-2017.pdf

[4] G. Yang and T. S. Huang, “Human face detection in

complex background,” Pattern Recognition Letter, vol. 27,
no.1, pp. 53-63, 1994.

[5] C. Kotropoulos and I. Pitas, “Rule-based face detection in

frontal views,” Proc. Int’l Conf. Acoustics, Speech and Signal
Processing, vol. 4, pp. 2537-2540, 1997.

[6] Xinjun Ma, Hongqiao Zhang, XinZang, “A face detection

algorithm based on modified skin-color model”, CCC, vol. 1,
pp. 3896-3900, IEEE, 2013

[7] V. Shehu and A. Dika, “Using Real Time Computer

Algorithms in Automatic Attendance Management Systems.”
IEEE, pp. 397 – 402, Jun. 2010.

[8] Prof. P.K Biswas, Digital Image Processing

[9] M. A. Turk and A. P. Pentland, “Face Recognition Using

Eigenfaces”in Proc. IEEE Conference on Computer Vision and
Pattern Recognition, pp. 586–591. 1991

[10] W. Zhao, R. Chellappa, P. J. Phillips, and A. Rosenfeld,

“Face recognition: A literature survey,” ACM Computing
Surveys, 2003, vol. 35, no. 4, pp. 399-458

[11] Bhumika G. Bhatt, Zankhana H. Shah “Face Feature

Extraction Techniques: A Survey”, National Conference on
Recent Trends in Engineering & Technology, 13-14 May
2011.

[12] Adam Schmidt, Andrzej Kasinski, “The Performance of

the Haar Cascade Classifiers Applied to the Face and Eyes
Detection”, Computer Recognition Systems 2

© 2018, IRJET | Impact Factor value: 6.171 | ISO 9001:2008 Certified Journal | Page 4610