Face Recognition Attendance System
Tathagata Rudra
University Roll No. 12017009001255
Dibyadarshi Kashyap
University Roll No. 12017010001001
Rohit Ghosh
University Roll No. 12017009001046
Srishti Kumari
University Roll No. 12017009001463
Anjali Kumari
University Roll No. 12017009001418
Arunava Kundu
University Roll No. 12017009001260
Introduction
► Face recognition has taken a dramatic change in today’s world of, it has
been widely spread throughout last few years in drastic way. Their have
been some drastic improvements in last few years which has made it so
much popular that now it is being widely used for commercial purpose as
well as security purpose also. Tracking a users presence is becoming one
of the problems in today’s world.
► An attendance system based on facial recognition can act as a real world
solution to this problem and add great heights of simplicity for tracking a
users attendance. The manual entering of attendance in logbooks
becomes difficult and takes a lot of time also, so we have designed an
efficient module that comprises of face recognition using LBPH
algorithm(Open CV) to manage the attendance records of employee or
students.
Introduction
► During enrolling of a user, we take multiple
images of a user along with his/her id/roll
number and name also. The presence of each
student/employee will be updated in database.
► The results showed improved performance over
manual attendance system. This process can give
us more accurate results in user interactive
manner rather than the existing attendance
systems.
Literature Survey:
► Arun Katara et al. (2017) mentioned disadvantages of RFID (Radio Frequency Identification) card system, fingerprint system
and iris recognition system. RFID card system is implemented due to its simplicity. However, the user tends to help their
friends to check in as long as they have their friend’s ID card. The fingerprint system is indeed effective but not efficient
because it takes time for the verification process so the user has to line up and perform the verification one by one.
However for face recognition, the human face is always exposed and contain less information compared to iris. Iris
recognition system which contains more detail might invade the privacy of the user. Voice recognition is available, but it is
less accurate compared to other methods. Hence, face recognition system is suggested to be implemented in the student
attendance system.
► Difference between face detection and face recognition are often misunderstood. Face detection is to determine only the
face segment or face region from image, whereas face recognition is to identify the owner of the facial image.
S.Aanjanadevi et al. (2017) and Wei-Lun Chao (2007) presented a few factors which cause face detection and face
recognition to encounter difficulties. These factors consist of background, illumination, pose, expression, occlusion,
rotation, scaling and translation.
► Akshara Jadhav et al. (2017) and by P. Arun Mozhi Devan et al. (2017) suggested Viola-Jones algorithm for face detection
for student attendance system. They concluded that out of methods such as face geometry based methods, Feature
Invariant methods and Machine learning based methods, Viola-Jones algorithm is not only fast and robust, but gives high
detection rate and perform better in different lighting condition.
► There are a few feature extraction methods for face recognition. The original LBP (Local Binary Patterns) operator was
introduced by the paper of Timo Ojala et al. (2002). In the paper by Md. Abdur Rahim et al. (2013), they proposed LBP to
extract both texture details and contour to represent facial images. LBP divides each facial image into smaller regions and
histogram of each region is extracted.
Current Situation Where Is This
Presently In Use:
► Face is a very crucial part of human body, so we are using this to identify a
person using face biometric
► We use this method to recognise face in real time from image videos.
► Beneficial to “ LAW enforcement and crime branch agencies”.
► Surveys.
► Forces to find the missing or dead one’s.
► Security purposes.
► Helping the blinds.
► Marking down attendance.
Flow Diagram Of The Whole Process
The Camera first incorporates a
input image of the user and
saves them in the directory.
Then the input dataset is been
fed as input to the trained
system where the trained file
been created with the data
taken.
Now if the user again comes in
front then the face is identified
and the corresponding data is
being sent to the database and
the attendance of the authorised
user is marked on webpage
record.
Face Detection
► The objective of detection is to detect and distinguish a particular type of
subject from an image. To do this, we use Haar-Cascade Classifier using
OpenCV package.
► Haar Cascade Classifier is an ML Algorithm that aids object detection from
an image or frame of videos, based on concepts of feature in "Rapid Object
detection using a Boosted cascade of Simple features- 2001" by Paul Viola
and Michael Jones.
► Provide a lot of images of the object of interest as positive images and other
arbitrary images as negative images
► Then, the task is to get the features from the images.
Haar-Cascade Classifier
► First, Haar-like features are extracted, which help in understanding the
images.
► Then to make calculation faster, we use Integral Images.
► Then we make multiple weak classifiers, such that each classifier
attempts to rectify the error of previous classifier. As we keep adding
the weak classifiers we get a strong classifier. This is called Ensemble
► The final step is to cascade. In this the idea is to look for features in
each sub-window, if features not present, it skips the sub-window and
move on to the next one.
Face Recognition
► Feature Extraction:
Different Facial images means there are changes in
textural or geometric information.
In order to perform face recognition, these features have
to extracted from the facial images and classified
properly.
OpenCV Currently supports 3 Face Recognizer Algorithms,
EigenFaces, FischerFaces and Local Binary Pattern
Histograms. For Our Purpose, we are using the last one
LBPH Algorithm.
Choosing LBPH Over other 2 Face Recognizer Algorithm
plainly because of accuracy reasons and on the fact how
these algorithms performs on various illumination
scenarios, it is found that Accuracy of EigenFaces and
FischerFaces stands at 77.97% and 82.45% whereas LBPH
Stands at 90.93%.
LBPH extracts the local grayscale features by performing
feature extraction on a small region throughout the
entire image.
Working Principle Of LBPH
► The basic idea of Local Binary Patterns is to summarize the local structure in
an image by comparing each pixel as center and threshold its neighbours
against.
► If the intensity of the center pixel is greater-equal to its neighbour, then it
should be denoted with 1 and 0 if not. We’ll end up with a binary number for
each pixel, just like 11001111. So with 8 Surrounding pixels we’ll end up with
2^8 possible combinations, called Local Binary Patterns.
► The LBP Operator in a fixed 3x3 neighbourhood looks like this:
► LBP Operator is robust against monotonic grayscale transformations. We can
verify this by looking at the LBP image of an artificially modified image.
► After that what's left to do is how to incorporate the spatial information in
the face recognition model. The LBP Image is divided into m local regions and
extracts a histogram from each region. The spatially enhanced feature vector
is obtained by concatenating all the local histograms and this feature vector
is used to recognise unique faces. These Histograms are called Local Binary
Pattern Histograms.
► OpenCV supports LBPH Algorithm and we can use its functionalities if we
import OpenCV package.
//recognizer= cv2.face.LBPHFaceRecognizer_create()
Advantages of using facial recognition
for attendance system
►Increased security
►Automated time tracking system
►Time saving
►Easy to manage
►Cost-effective
Disadvantages of using facial
recognition for attendance system
►Data privacy breach
►If in the wrong hands, it will be a disaster
►Low reliability
►Provide opportunities for fraud and other crimes
►Lack of regulations
Scope Of Improvement
► In this proposed approach, there are a few limitations. First, the input image has
to be frontal and a upright single facial image. Second, the accuracy might drop
under extreme illumination problem. Third, false recognition might occur if the
captured image is blurred. Besides, LBP is textural based descriptor which extracts
local features. Hence, test image and train image have to be the same quality
which is captured by using the same device in order to have high accuracy.
► For future work, a better camera and a better lighting source can be used in order
to obtain better result.
► For Multiple Face Detection and more accurate recognition purposes we can
implement A.I (artificial intelligence) algorithm to perform the face recognition.
CNN (Convolution Neural Network) which is a hot topic recently, is a machine deep
learning algorithm which is able to perform recognition with less dependency on a
particular train image given a large database. However, CNN requires an extremely
large database to increase its accuracy or having relatively small class size to have
high performance.
Future Scope
► This system helps to avoid fail proof of attendance system and acts as a
substitute of existing systems e.g. Biometric Attendance System.
► This system takes less time and much efficiently records attendance and
hence considered as time saving and secured attendance system.
► Further extensions can be made to achieve real time face detection in
surveillance premises and instead of taking images we can work with videos.
► The future work is to improve the recognition rate of algorithm if we want to
achieve real time face detection through surveillance cameras.
► The system should be improved further so that it recognizes face with more
than 30 degrees of angle variation.
Conclusion
► The development of this technology has not only ensure attendance to be
taken accurately but also eliminate the flaws in the previous systems.
► Using technology for maintaining attendance system will reduce human
intervention in the whole process by handling all the task through machine.
► The only cost to this solution is to have sufficient storage space to store all the
faces in the storage. Fortunately we have enough storage devices available to
compensate with the large amount of data.
► At the end, the system not only resolve troubles that exist in the old model but
also provide convenience to the user to access the information collected which
perfected the existence of technology to assist human’s needs.
References:
► Cascade Classifier OpenCV:
https://docs.opencv.org/3.4/db/d28/tutorial_cascade_classifier.html
► Ahonen, Timo, Abdenour Hadid, and Matti Pietikainen. “Face description
with local binary patterns: Application to face recognition.” IEEE
transactions on pattern analysis and machine intelligence 28.12 (2006):
2037–2041.
► LBPH OpenCV:
https://docs.opencv.org/2.4/modules/contrib/doc/facerec/facerec_tutoria
l.html#local-binarypatterns-histograms
► Local Binary Patterns: http://
www.scholarpedia.org/article/Local_Binary_Patterns
