© 2018 IJSRST | Volume 4 | Issue 5 | Print ISSN: 2395-6011 | Online ISSN: 2395-602X

Themed Section: Science and Technology

Implementation of Smart Attendance System Using Raspberry

Pi
Rashmi R. Rumane1, Dr. Ajay. A. Gurjar2
1M.E Student, Department of Electronics & Communication Engineering, Sipna College of Engineering and
Technology, Amravati, Maharashtra, India
Professor, Department of Electronics & Communication Engineering, Sipna College of Engineering and
2

Technology, Amravati, Maharashtra, India

ABSTRACT

Being one of the most successful applications of the image processing, face detection and recognition has a vital
role in technical field especially in the field of security purpose. Human face detection and recognition is an
important field for verification purpose especially in the case of attendance system. Maintaining the attendance
is very important in all the institutes for checking the presence of students. Every institute has its own method
in this regard. Some are taking attendance manually using the traditional pen and paper or file based approach.
This system is developed for deploying an easy and a secure way of taking down attendance. The system first
captures an image of all the students and stores the information into database. The system then stores the image
by mapping it into a face coordinate structure. Next time whenever the registered student enters the premises
the system recognizes the student and marks his attendance along with the time. In this project, we come up
with a new hardware system for human face detection which makes use of Raspberry Pi. It is a credit-card
sized computer with the components mounted on a credit card sized motherboard, running a dedicated version
of Linux. It plugs into TV and a keyboard. It is a capable little computer which can be used in electronic devices
and for much functionality that a desktop computer can perform. It comes at a very low price
Keywords: Raspberry pi, Face Detection, Face Recognition, Attendance System, Linux (Python)

I. INTRODUCTION losses, we are about to use automatic process which is

based on raspberry pi. In this novel approach, we are
Organizations of all sizes use attendance systems to using face detection & face recognition system. This
record when student or employees start and stop work face detection differentiates faces from non-faces and
and the department where the work is performed. is therefore essential for accurate attendance. The
Some organizations also keep detailed records of other strategy involves face recognition for marking
attendance issues such as who calls in sick and who the student‟s attendance. The Raspberry pi module is
comes in late. An attendance system provides many used for face detection & recognition. The camera will
benefits to organizations. Traditional approach for be connected to the Raspberry pi module. The
attendance is professor calls student name & record database is collected. The database includes name of
attendance. It takes some time to record attendance. the person, their images . One of the unique features
Suppose duration of class of one subject is about 50 of our brain is that it can think only in images not in
minutes & to record attendance takes 5 to 10 minutes. words. Once you may forget to keep your Car‟s key
For each lecture this is wastage of time. To avoid these but you will never forget to bring a face with you.

IJSRST1738260 | Received : 05 April 2018 | Accepted : 25 April 2018 | March-April-2018 [ (4) 5 : 1650-1658 ] 1650-
165
 Rashmi R. Rumane et al. Int J S Res Sci. Tech. 2018 Mar-Apr;4(5) : 1650-1658

God has given everyone a unique face. Face is the implemented using a non-intrusive digital camera
most important part of our body, so that it can reflect installed on a classroom, which scans the room,
many emotions of a person. There are traditional ways detects and extracts all faces from the acquired
are being utilized even today but with vast resources images. After faces have been extracted, they are
wanted to be more secured. There are two types of compared with an existing database of student images
biometric as physiological characteristics (face, and upon successful recognition a student attendance
fingerprint, finger geometry, hand geometry, palm, list is generated and saved on a database. This paper
iris, ear and voice) and behavioral characteristics (gait, addresses problems such as real time face detection
signature and keystroke dynamics). Sometimes your on environments with multiple objects, face
behavioral traits may change because of illness, fear, recognition algorithms as well as social and
hunger etc. Face detection and recognition system is pedagogical issues with the applied techniques.
more expensive, exact, easy to understand and non
intrusive process as compare to other biometrics. The In [2] PAN Xiang described work process of a system:
system will fall into two categories as face detection When a person wants to enter the access control
(1:1) and face recognition (1:N). In the face detection system, he used the RFID card to swiping card by
we have to classify between face versus non face non-touch way. The system reads the information in
region while in recognition process we have to the card and meanwhile the video camera is started to
compare that single face image with multiple images take photos of the person. Then the face can be
from the input image. detected in a short time. The identity information in
While capturing an images from a webcam we have to the card is compared to the information from the
come across some problems like pose (position of database and the corresponding face data will be
camera), presence of structural components obtained. If the identity information and the face data
(spectacles and beard), facial expression, occlusion are all matched to the information from the database,
(obstructed by someone), image orientation (variation the person will be passed. Else he can‟t enter. The
in rotation), imaging condition (lightning and camera manager can do the manage work such as query the
characteristics) etc. records.
A Face Detection and Face recognition System is a
system which automatically identifies and/or verifies In [3] Mr. Jawale described a technique based on ear is
the identity of a person from digital images or a video also introduced that is a photo of the subject‟s ear is
frame from a video source. A general statement of the taken and fed into the computer. Edge detection is
face recognition problem (in computer vision) can be carried out on this picture. From this detected edge, is
formulated as follows: Given still or video images of a separated a reference line with respect to which other
scene, identify or verify one or more persons in the features are identified. These extracted features are
scene using a stored database of faces. stored in a database in the form of a vector, each
vector corresponding to a particular image in the
II. LITERATURE REVIEW database. The feature vector of the test image obtained
is compared with those in the vector database, For
Two researchers Visar Shehu and Agni Dika creating and maintaining database for records of
proposed in [1] a system which introduces an individuals and feature vectors, which are used for the
attendance marking system, which integrates purpose of comparison and decision making, linking
computer vision and face recognition algorithms into of MATLAB and some data base using ODBC Drivers
the process of attendance management. The system is is carried out according to which a match is calculated.

International Journal of Scientific Research in Science and Technology (www.ijsrst.com) 1651

 Rashmi R. Rumane et al. Int J S Res Sci. Tech. 2018 Mar-Apr;4(5) : 1650-1658

This match is compared with a predecided threshold

value, which decides the identity of the person.

The [4] Jian Xiao, Gugang Gao, Chen Hu, Haidong

Feng proposed a framework for fast embedded face
detection system based on three modules. One fast
face detection method based on optimized AdaBoost
algorithm with high speed and high detection rate,
one SOC hardware framework to speed up detection
operations and one software distribution strategy to
optimize the memory sub-system.

The [6] describes a Real Time System developed for Figure 3.1: Proposed work diagram of Face Detection
Multi-face detection. As most of the system are based and Recognition
on software algorithms. This proposed system is based
on hardware design to enhance the processing time. The proposed work follows:
The different stages of this hardware design includes
skin color detection, morphology, Fast connected- The first part was booting the Raspberry Pi board by
component labeling algorithm, Implementation of the installing the Operating system Raspbian OS and
Fast connected-component labeling algorithm, Lip installing the essential libraries and packages.
feature extraction, Horizontal edge detection.
1) Interfacing of camera module with raspberry pi
III. PROPOSED WORK B3 model. Capture face image using USB
Webcam i.e. image acquisition take place
3.1 SYSTEM DESIGN 2) Create a database of authorized person using face
The aim of our dissertation is to provide a attendances detection process. After created database save
system using face detection and recognition on into folder.
Raspberry Pi board and send an alert message to the 3) Capture current face, compare with data base
authorized person via mobile device or internet image. Using face detection and face recognition
server This will increase the security of our process. Algorithms are used for face detection
Dissertation. Whatever the system that we have to and recognition.
implement it should work in real time with low false 4) If any person is not present then message will
detection rate. Figure. 1 shows the block diagram for be send to absent person.
real time face detection and recognition system that
will contain various hardware and software
components.

International Journal of Scientific Research in Science and Technology (www.ijsrst.com) 1652

 Rashmi R. Rumane et al. Int J S Res Sci. Tech. 2018 Mar-Apr;4(5) : 1650-1658

3.2 SYSTEM ARCHITECTURE

Figure 3.3: USB Webcam

3.2.2 Raspberry pi B3 Module

Figure 3.2: Architectural block diagram of

Implementation of human face detection and
recognition system for attendances

WORKING
Above figure show the basic block diagram our
dissertation system can be operated in two different
sections, i.e. one for capturing and creating a data base
and the other section is to capture the image and
which is used for identifying or comparing the images
in the database for attendances mechanism.

3.2.1 USB Webcam

The imaging module in the proposed system is
realized using a USB webcam.USB Webcam connects
to the raspberry pi b3 module. Its resolution is Fig3.4: Raspberry pi B3 Model
8megapixel and still picture resolution 640x480, Max
image transfer rate 1080p: 30fps. USB Webcam The Raspberry Pi 3 is the third generation Raspberry
module is used for captures an image and send Pi. The Raspberry Pi 3 uses a Broadcom BCM2837 SoC
captured image to the Raspberry pi module. At the with a 1.2 GHz 64-bit quad-core ARM Cortex-A53
first time camera module captures images to create a processor, with 512 KB shared L2 cacheThe Raspberry
database of the authorized person as a train faces then Pi is a low cost, credit card sized single board
secondly take a test face or live captured image to computer developed by raspberry pi foundation.
compare with train faces (Data base). Raspberry pi is controlled by a modified version of
Debian Linux optimized for the ARM architecture.
The Raspbian OS distribution of Debian is designed to
provide embedded software programmers with a
familiar and fully functional interface to custom
hardware development. Raspberry pi consist of OPEN

International Journal of Scientific Research in Science and Technology (www.ijsrst.com) 1653

 Rashmi R. Rumane et al. Int J S Res Sci. Tech. 2018 Mar-Apr;4(5) : 1650-1658

CV library in which we can write algorithm in JAVA, Step 1: Setting up Raspberry PI

Python.. SD card into the SD card slot on the Raspberry Pi,
Raspberry pi B3 module is small computer board. which will only fit one way. Next, plug the USB
Webcam interface with raspberry pi model this keyboard and USB mouse into the USB slots on the
capture face image fed to the raspberry pi model. Raspberry Pi. Make sure that the monitor or TV is
Raspberry pi contain SD card i.e. memory card face turned on, and that the right input is selected. Then
image of database authorized person stored on this. W Connect the HDMI cable from your Raspberry Pi to
captures face image compare with database image. the monitor or TV. If the Raspberry Pi is needed to be
Face not recognized with database then send message connected to the internet, plug an Ethernet cable into
to person mobile. the Ethernet port next to the USB ports. When all the
required cables and SD card are plugged then, plug in
IV. IMPLEMENTATION the micro USB power supply. First of all the Linux OS
Below figure shows that the flowchart of is installed into the Raspberry Pi board via Micro SD
implementation of Human face detection and card and appropriate code is written in Python
recognition system using raspberry pi language for the object detection using the Open CV
libraries and is dumped in the board. The USB
Camera is interfaced; the GPIO pins are programmed
using commands in Linux and Python in this stage.
The camera is interfaced to the Pi via the USB port
and the door lock module is interfaced via the GPIO
pins on the Pi.

Step 2: Image Acquisition

Webcam connect to the raspberry pi kit. When
person standing in front of camera then webcam
capture face image of that person. Resize the captured
face image.

Step 3: Pre-processing
The data which is collected from Input unit that is
captured Image and Video frames input is fed into the
processing unit in the processing unit. Here the
processing unit is nothing but a Raspberry Pi board.
Along with code scripts of the implemented modules.

Step 4: Face detection

In face detection process haar cascade algorithms are
Figure 4.1 Flow chart of Implementation of Human used for face detection.
Face Detection and recognition System for  First capture face image by using webcam then
attendances system using Raspberry pi resize the image.
 Haar-like features are digital image features used
Following is implementation process of dissertation in object detection.

International Journal of Scientific Research in Science and Technology (www.ijsrst.com) 1654

 Rashmi R. Rumane et al. Int J S Res Sci. Tech. 2018 Mar-Apr;4(5) : 1650-1658

 Initially, the algorithm needs a lot of positive V. RESULT

images (images of faces) and negative images Five basic face recognition techniques are compared
(images without faces) to train the classifier. for different algorithms based on recognition rates
(percentage) and elapsed time is shown in table 5.1.
Then we need to extract features from it. In this face
The common training data set and test data set is used
detection process, captured colour face image is
for comparison in all four used techniques. In this
converted into black and white image i.e. gray scale
work, basically the comparison of various algorithms
image. It reduce the number of pixel for simplification
as stated above i. e; PCA, KPCA, LDA , LEM and LBP
of detection. Then selection of ROI take place i.e. face
is done on the basis of recognition time used in face
detection process is done.
recognition procedure.

Step 5: Create Data base of Authorized Person

Table No 5.1: Comparison of various methods on the
In Face detection process first create database of
basis of recognition rate and recognition time
authorized person. The captured face of current poses
creates a data base of the authorized person and stores
Sr Methods Recognition Recognition
this. At the next time camera module will capture the
No. Time (sec) Rate(%)
current live face of the person. All this process is done
in Raspberry pi module. Create our own database. The
images stored in a folder, storing database images and
store it as a train faces in the test 1 folder. 1. PCA 24.13 98.99

Step 6: Face Recognition 2. KPCA 20.33 94.99

The most important step of our dissertation is face
recognition. In Face recognition process LBPH
3. LDA 33.80 97.99
algorithms are used.
 The image is divided into small connected regions
called cells, and for the pixels within each cell, a 4. LBP 7.5 95.05
histogram of gradient directions is compiled.
 For improved accuracy, the local histograms can
be normalized by calculating a measure of the After executing all the step of implementation as
intensity across a larger region of the image, called discussed above we get following results as shown in
a block, and then using this value to normalize all figures given below:
cells within the block.
 The final step in object recognition using Local 5.1 database generation
binary pattern(LBP) algorithm. In our project the main task is to create a database of
students enrolled in a class and then to campare these
Step 7: Output images with the live captured images. These captured
When captured face match with database face image images help in identifying the known or unknown
i.e. face recognized successfully. When person are person.The face images get stored in a default folder.
absent then send absent message to person.

International Journal of Scientific Research in Science and Technology (www.ijsrst.com) 1655

 Rashmi R. Rumane et al. Int J S Res Sci. Tech. 2018 Mar-Apr;4(5) : 1650-1658

known person, therefore attendance of present

student is automatically marked.

5.3 face detection of unknown student

Figure.5.3 shows that the captured image is of
unknown person. Because we have created an
enrolled students database in the system and this
captured image is unknown to the system after
comparing with the database and hence indicated by
red colored rectangle.
Hence attendance can not be marked.

Figure 5.1: Created database

Fig.5.1 shows that database of enrolled student is

successfully generated and stored in default folder,

5.2 Face detection of enrolled student

Now live captured image is taken as a test face shown
in figure 5.2 to identify whether the student is
enrolled or not.

Fig 5.3:image captured of unknown person

5.4 Email to the head authority

Fig 5.2: face detection after entering a classroom

Fig 5.2 shows that captured imaged is detected and it

is available in the database and hence face is indicated
by blue coloured rectangle and also shows name of the

International Journal of Scientific Research in Science and Technology (www.ijsrst.com) 1656

 Rashmi R. Rumane et al. Int J S Res Sci. Tech. 2018 Mar-Apr;4(5) : 1650-1658

Fig 5.4: Email sent to the head authority

Fig 5.5: attendance sheet of present students

Fig 5.4 shows that attendance of present students

Fig 5.5 shows attendance sheet of students present for
during the lecture is recorded and sent to the
the lecture along with roll number and time.
authority along with the detected image and name.

VI. CONCLUSION
5.5 Attendance sheet
We have to study face detection and recognition
After completion of the lecture, attendance sheet is
system on Raspberry Pi module. Face detection and
generated and stored in default folder.
recognition is currently a very active research area.
This attendance sheet consists of name of student, roll
Some of the more algorithms are still too
number, date and time.
computationally cheap to be applicable for real time
processing. Other processors are costlier than
Raspberry Pi along with large memory, accuracy and
speed. Using Python and Open CV in Raspberry Pi,
made our project flexible. But in future it can be used
in Orange Pi and Banana Pi board. This has more
RAM.

VII. REFERENCES

[1] M. Molina Shamanth G S ashwin Kashyap „„face

detection using Raspberry pi and python,‟‟
.NCPD 2016 July 2016.
[2] S. C. Gaddam, N. V. K. Ramesh and Hema Dsn

International Journal of Scientific Research in Science and Technology (www.ijsrst.com) 1657

 Rashmi R. Rumane et al. Int J S Res Sci. Tech. 2018 Mar-Apr;4(5) : 1650-1658

[3] hanekula,‟‟ Face Recognition Based Attendance [13] Navin Prakash, Dr. Y. singh, ‟‟Support vector
Management System with Raspberry PI 2 using Machines for Face Recognition ‟‟IRJET Vol: 02
Eigen Faces Algorithm, ‟‟ARPN journal vol -11, issue: 08|Nov-2015.
NO.13, july 2016. [14] JunLee, Jeong-Sik Park, Gil-Jin Jang and Yong-
[4] Raspberry Pi Face Recognition Treasure Box Hoseo,‟‟ Efficient Head Pose Determination and
Created by Tony Di Cola. its Application to face Recognition on Multi-Pose
[5] MedakTeenaRavali, Prof. RangasaiKomaragiri face DB,‟‟IJMUE Vol.11.No2 (2016).
“Image Processing Platform On [15] G. Senthilkumar, K. Gopalakrishnan,
Raspberry Pi For Face Recognition” Global V.S.Kumar,‟‟ Embedded Image Capturing System
Journal of Advanced Technologies, ISSN 2277- Using Raspberry PI System,‟‟ IJETTCS Volume 3,
6370 Vol3, Issue4- 2014. Issue 2, March-April 2014.
[6] R.Pahune ,A.A.Chaudhuri,”Face Detection [16] Gheorghita Ghinea ,Rajkumar Kannan and
System for Security Purpose Using Raspberry Suresh Kannaiyan, ”Gradient orientation-based
PI,”ICEIS-2016. PCA Subspace for Novel Face Recognition ”in
[7] Richard Mo Adnan Shaout, ‟‟Portable Facial IEEE Access, Vol 2,2169-3536 2014.
Recognition Jukebox using Fisher Faces, [17] Mounika B.R Reddy N. Jand Reddy V.B.D,‟‟A
‟‟IJACSA Vol.7, No.3, 2016. Neural Network Based Face Detection Using
[8] Subhanishaik, Antomicheal,‟‟Automatic Age and Gabor Filter Response, ‟‟IJNN ISSN:2249 -2763
Gender Recognition in Human face Image Data &ESSN: 2249-2771, volume 2 ,Issue 1, 2012 ,PP.-
Set Using Convolutional Neural Network 06-09.
System,‟‟IJARCSMS Vol 4, issue 2, Feb 2016 [18] Tudor Barbu,‟‟Gabor Filter –Based Face
.ISSN:2321-7782. Recognition Technique, ‟‟Processing of the
[9] Danwei, ‟‟Distributed Compressive Sensing Based Domain Academy, Series A, Vol 11, No 31
Near Infrared and Visible Images Fusion for Face 2010,PP.277-283.
Recognition,‟‟ IJSP, IPPR vol.9, No.4 (2016), [19] H.R.Kanan and K.Faez,‟‟Adaptively Weighted
pp.281-292. Sub –Gabor Array for Single Model –based
[10] Unsoojang and Eui Chul lee, ‟‟Performance human face Recognition‟‟. IJECE DEC 2010.
Analysis of Pixel Based Face Recognition [20] P. Shih and, C. Liu, ‟‟Face Detection using
Method, ‟‟IJBSBT Vol.8, No. 2(2016), PP.197-206. Discriminating Feature Analysis and Support
[11] Li yongqiang and panjin, „„one sample image Vector Machine, ‟‟.Pattern Recognition Vol.39,
Recognition algorithm based on improved sub- NO.2, Feb 2006, PP.260.276.
pattern principle component analysis, ‟‟IJSP,IPPR [21] S.K.Nabeena. N. V. Narayana Rao,‟‟
vol.8,NO.9(2015),PP.77-84. Implementation of Video Surveillance Using
[12] Maryam Moghaddam, Saeed Raspberry PI 2,‟‟ IJMETMR ISSN No:2348-4845.
Meshgini,”Automatic face Recognition via Local [22] Sumo .M.O, Rashmi .H.N, S.B.Seshadri.‟‟ Stand
Directional Patterns,”journal of Artificial Alone Face Recognition System Using Principle
Intelligence in Electrical Engineering, vol.4, Component Analysis,‟‟ IJETCAS.
NO.15, DEC 2015.

International Journal of Scientific Research in Science and Technology (www.ijsrst.com) 1658