Volume 7, Issue 7, July – 2022 International Journal of Innovative Science and Research Technology

ISSN No:-2456-2165

“University for Industry”

ADDIS ABABA SCIENCE AND TECHONOLOGY UNIVERSITY
College of Electrical and Mechanical Engineering
Departments of Electrical and Computer Engineering
(Computer Engineering Stream)
BSC. Final Year Project

Computer Vision Based Authentication and

Employee Monitoring System
1. Getachew Getu ………………… ETS0557/10
2. Yinebeb Tariku …………………ETS1263/10
3. Yosef Emyayu …………………ETS1310/10

Advisor name: Mr. Ashenafi Yadessa

June 20, 2022

IJISRT22JUL1344 www.ijisrt.com 827

Volume 7, Issue 7, July – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
EXAMINING COMMITTEE APPROVAL SHEET

TITLE OF THESIS:
Computer Vision Based Authentication and Employee monitoring System

Student Names (ID)

Getachew Getu (ETS0557/10)
Yinebeb Tariku (ETS1263/10)
Yosef Emyayu (ETS1310/10)

Approved by the examining committee members:

NameAcademic-Rank Signature Date
Advisor: _____________ __________________ ____________
Examiner:____________ _________ _____________________
Examiner:____________ _________ _____________________

Name SignatureDate
DC Chairperson: ____________________ _____________ ____________
Associate Dean for
Under Graduate Program_________________ ______________ __________

IJISRT22JUL1344 www.ijisrt.com 828

Volume 7, Issue 7, July – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165

DECLARATION

We, hereby declare that, this paper titled “Computer vision-based authentication and employee
monitoring system” bases on the results that we have derived ourselves throughout our research. Materials
of work from researches conducted by others are mentioned in the references. Finally, it is a must to tell you
how much effort we have made to do this project and come up with tangible outcomes.

Declared by (students): Signature Date

1. Getachew Getu _______________
2. Yinebeb Tariku _______________
3. Yosef Emyayu _______________

Approved by:
Advisor name: ________________________
Signature: ___________________________

IJISRT22JUL1344 www.ijisrt.com 829

Volume 7, Issue 7, July – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165

ACKNOWLEDGMENT
First, we are glad to thank the almighty GOD for giving us the potential to do this project. Next, our
advisor Mr. Ashenafi Yadesa, he guides, supportand continuous follow up our effort to come up with these
points. It was his perpetual motivation and guidance that has helped us to carry out this project. We also
want to thank Mr. Melaku who is a computer engineering lab assistant in AASTU and he supports us by
allowing a lab room and by giving some electronics components in order to achieve this project.

Last but not least we would like to thank Prof. Ramasamy for his willingness to allow us use a computer
vision lab and the equipment we need.

IJISRT22JUL1344 www.ijisrt.com 830

Volume 7, Issue 7, July – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165

TABLE OF CONTENT

Declaration ............................................................................................................................ 829

Acknowledgment .................................................................................................................. 830
Abstract ................................................................................................................................. 833
List of figures ........................................................................................................................ 834
List of tables .......................................................................................................................... 834
List of Abbreviation and Acronym ...................................................................................... 835
Chapter One.......................................................................................................................... 836
Introduction .......................................................................................................................... 836
1.1 Background Study .................................................................................................. 836
1.2 Statements of the Problem ..................................................................................... 837
1.3 Objective ................................................................................................................. 837
1.3.1 General objective ............................................................................................. 837
1.3.2 Specific objective ............................................................................................. 837
1.4 Scope of the project ................................................................................................ 837
1.5 Literature review .................................................................................................... 837
1.6 Methodology ........................................................................................................... 838
Chapter Two ......................................................................................................................... 840
System Requirements, Architectural Design and Specifications ........................................ 840
2.1. System requirements .............................................................................................. 840
2.1.1. Functional requirements ................................................................................. 840
2.1.2. Non – functional requirements........................................................................ 840
2.1.3. Hardware and Software requirements ........................................................... 840
2.2. Architectural Design and Specifications ................................................................ 841
2.2.1. Introduction ..................................................................................................... 841
2.2.2. Systems (Raspberry PI) detail architecture.................................................... 842
Chapter Three....................................................................................................................... 844
Engineering Design and Detail Specification ....................................................................... 844
3.1 Engineering design and detail specification ........................................................... 844
3.2 Algorithm Design .................................................................................................... 851
3.3 System’s hardware/ circuit configuration(implementation) ................................. 852
3.4 Mechanical Design .................................................................................................. 853
Chapter Four ........................................................................................................................ 857
Implementation, testing and Discussion .............................................................................. 857
4.1 Implementation ....................................................................................................... 857
4.1.1 Component connection through GPIO pins ................................................... 858
4.1.2 GUI and python Scripts uses ........................................................................... 859

IJISRT22JUL1344 www.ijisrt.com 831

Volume 7, Issue 7, July – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
4.1.3 Detail code explanation (implementation) ...................................................... 860
4.2 Testing and Discussion ........................................................................................... 867
Chapter five .......................................................................................................................... 871
Conclusion and future works ............................................................................................... 871
5.1 Conclusion............................................................................................................... 871
5.2 Future work ............................................................................................................ 871
References ............................................................................................................................. 872
Appendix ............................................................................................................................... 872

IJISRT22JUL1344 www.ijisrt.com 832

Volume 7, Issue 7, July – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
ABSTRACT

Computer vision is a recently active topic which is based on image processing. Computer vision
used in variety of topics from a simple feature detection up to a sophisticated system those need
animage as in modern robots. In this project, we are going to use computer vision to make an
authentication system for employee of a company to have an access to the campus or for specific
resource like entering to a datacenter. By using the art-of-state of CV, we develop the system which
screen users by using their face which is more secured and have no doubt of repetition since facial
features are unique than other methods such as password authentications, those are less secured and
vulnerable to attack.

Existing python built in database called SQLite is used inmanage employee information such as
their ID number, full name, department within a database. We deploy Raspberry PI as a main
controller/ system on board computer. We have developed an interactive GUI using python’s
librarytkinter, for the admins to manage their task those are registering new employee, train the
model over the new added employee’s faces, and monitoring the system overall functionality. Every
employee needs to be registered via the admin’s monitoring GUI and once their faces are added and
trained on the system/model, the system will process either to give permission or deny request for an
entrance viaa servomotor which actuate the opening and closing of a door.Previously unregistered
employee can ask a request to access the resource via guest mode which will processed and respond by
the respective admins.

Since CV based face detection is challenged by light intensity easily and need an advanced
calibration to overcome the effect, we develop the two-factor authentication. User need to provide
his/her password, then the system generates a one-time password which is time bounded and the
communication will process via email. Lastly after we designed and analyze the high-level system,we
implement the whole system’s prototype with available resources.

Keywords:- Computer vision (CV), Face recognition, OpenCV, and OTP.

IJISRT22JUL1344 www.ijisrt.com 833

Volume 7, Issue 7, July – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
LIST OF FIGURES

Figure 1.1 Procedures we follow in the project ....................................................................... 838

Figure 1.2 flow of operation.................................................................................................... 839
Figure 2.1 General system description. ................................................................................... 841
Figure 2.2 Our system description - on the rough .................................................................... 842
Figure 2.3 signal flow graph ................................................................................................... 843
Figure 3.1 raspberry pi and its pin configuration ..................................................................... 845
Figure 3.2 Liquid crystal display (16x2 LCD) ......................................................................... 845
Figure 3.3 One channel relay modules .................................................................................... 846
Figure 3.4 a typical SG90 servo motor .................................................................................... 847
Figure 3.5 camera module and its pin configuration ................................................................ 848
Figure 3.6 a. Heat source movement b. output signal ........................... 848
Figure 3.7 4x4 alphanumeric keypad and its internal structure ................................................ 849
Figure 3.8 communication between the local device raspberry pi and the remote server via network 849
Figure 3.9 flow chart for registration and training ................................................................... 851
Figure 3.10 flow chart for face-recognition ............................................................................. 852
Figure 3.11 System’s circuit connection (drawn in fritzing). ................................................... 853
Figure 3.12 mechanical design of the door system .................................................................. 853
Figure 3.13 the mechanical design for controlling door movement .......................................... 854
Figure 3.14 initial position of the door (door is closed) ........................................................... 854
Figure 3.15 the sliding door start to open ................................................................................ 855
Figure 3.16 the door reaches final position (opened) ............................................................... 855
Figur 3.17 initial to final movement of the firist link ............................................................... 855
Figure 4.1. Connection of Keypad, LCD, and PI camera with the Raspberry Pi. [a) drawn on fritzing, b)
image taken by our phone camera.] ......................................................................................... 866
Figure 4.2. admin login page................................................................................................... 866
Figure 4.3. Home windows ..................................................................................................... 867
Figure 4.4 new employee registration window ........................................................................ 867
Figure 4.5 training the model .................................................................................................. 868
Figure 4.6 Person face recognized, single person(a) and two people(b) ................................... 868
Figure 4.7 Email verification (OTP) during fog weather condition .......................................... 869
Figure 4.8 Guest mode email notification................................................................................ 869
Figure 4.9 admin remove the user ........................................................................................... 870

LIST OF TABLES

Table 3.1 hardware specification fort the design-level ............................................................. 850

Table 3.2 Hardware specification (for the implementation-prototype). .................................... 850
Table 4.1 component pin connection ....................................................................................... 859

IJISRT22JUL1344 www.ijisrt.com 834

Volume 7, Issue 7, July – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165

LIST OF ABBREVIATION AND ACRONYM

CV ……………………… Computer Vision

OTP ………………….… One Time Password
LBPH…………………. Local Binary Pattern Histogram
LCD ……………………... Liquid Crystal Display
OS …………………….... Operating System
GPU ……………………...Graphics Processing Unit
GUI ……………………... Graphical User Interface
DC …………………….... Direct Current
GPIO ……………………. General-Purpose Input/Output
USB ………………….…. Universal Serial Bus
RAM ……………………. Random Access Memory
VGI ……………………... Virtual Graphics Interface
DFG ……………………...Data Flow Graph
CSI ………………….…... Camera Serial Interface
PWM …………………...Pulse Width Modulation
SMTP …………………... Simple Mail Transfer Protocol

IJISRT22JUL1344 www.ijisrt.com 835

Volume 7, Issue 7, July – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165

CHAPTER ONE

INTRODUCTION
A. Background Study
Computer vision(CV) is a field of computer science that deals with replicating the complex parts of the
human visual system and getting the machines to comprehend and understand the visual details present in
the data[1]. CV is a very popular field that is used in almost all the domains that work with images and
videos. It finds itself solving really complex problems in a simple manner and it is easy to train and work
with as well. Computer vision enables computers and systems to derive meaningful information from digital
images, videos and other visual inputs and take actions or make recommendations based on that
information[1]. Face detection and recognition are the most important application in computer vision.

Face recognition technology is a biometric technology, which is based on the identification of facial
features of a person. Humans can recognize visual patterns all the time, and we obtain visual information
through our eyes. For a computer, whether it is a picture or a video, it is a matrix of many pixels. The
machine should find out what concept a certain part of the data represents within the data. For face
recognition, it is necessary to distinguish who the face belongs to in the part of the data that the machines
think of the face.

Face detection is the process of simply detecting the presence of a face in an image or video stream.
Face detection algorithm is to find out the coordinate system of all faces in one image. Face
recognition takes the faces detected from the localization phase and attempts to identify whom the face
belongs to. Face detection and recognition technology can be applied to various fields such as security,
biometrics, law enforcement, entertainment and personal safety to provide surveillance and tracking of
people[2].

The methods used in face detection can be knowledge-based, feature-based, template matching or
appearance-based. Each has advantages and disadvantages:
 Knowledge-based, or rule-based methodsdescribe a face based on rules. The challenge of this approach is
the difficulty of coming up with well-defined rules.
 Feature invariant methodswhich use features such as a person's eyes or nose to detect a face. It can be
negatively affected by noise and light.
 Template-matching methods are based on comparing images with standard face patterns or features that
have been stored previously and correlating the two to detect a face. Unfortunately, these methods do not
address variations in pose, scale and shape.
 Appearance-based methods employ statistical analysis and machine learning to find the relevant
characteristics of face images.

In today’s world by using smart devices, we aremaking our needssmart. By following trends and
updates we have to consider andremove drawbacks in existing system and add more featuresand updates.
The study of Open-CV and its inbuilt library functions helps to generate a code thatwill do correct and
authentic facial recognition system with new and more efficient use of hardware.

OpenCV (Open-Source Computer Vision) is a popular computer vision library started by Intel in 1999.
The cross-platform library sets its focus on real-time image processing and includes patent free
implementations of the latest computer vision algorithms[3].

Human body willbe identified as an input within environment by capturing livevideo from a camera and
the process will be done on captured video frames. The images will run through a processor such as
raspberry pi and check with the stored database.

IJISRT22JUL1344 www.ijisrt.com 836

Volume 7, Issue 7, July – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
B. Statements of the Problem
Strong and multifactor authentication system are needed at the current situation in many sectors. we
cannot trust the conventional methods namelya manual lock at the door, password keylock can be guessed or
observed and token card can be stolen. To solve the problems, we tried to make things better and efficient.
Face detection and face recognition with multifactor password authentication via computer vision is the
most advance and it will provide a better authentication feature than the conventional access.

The existing attendance system requires employees to manually sign the sheet every time they attend
their work. This adds more extra time to be consumed by the employee to find their name on sheet, some
employees may mistakenly signed other’s name and sometimes sheet may get lost. To solve the problem in
manual attendance system, we used face recognition as an authentication mechanism. After having images
from a camera and once recognized, the door would be open and the attendance would be recorded.

C. Objective
a) General objective
The main objective of this project is to provide multi factor authentication and employee monitoring
system by using computer vision.

b) Specific objective
 To design efficient face recognition system.
 To utilize the face recognition system based on OpenCV.
 To design and developmultifactor password authentication system.
 To generate and record user’s data who have been access the system.
 Develop an effective two-factor authentication via email.
 To design highly reliable and efficient and automated door opening-closing mechanism.

D. Scope of the project

In this system, we have the following in-scopes such as the authentication and employee monitoring
system with facial recognition-based attendance and authentication. For fog and smog weather conditions,
we add a two-factor authentication and a guest mode also included for unregistered users. For the project
implementation, we limit the scope to work only for local connection between admin and the system. This
project is applicable in home’s (rooms) door, data center and offices by using face recognition systems by
providing more security.

The remote server connection between the admins and the system is considered as out-scope of the
project. The case where employee may leave the staff once they take attendance is out of the scope. Also, a
possible entrance of a person other than the registered person at time of door opening also out of the scope.

E. Literature review
We have discovered various papers related with the topic. The author in [4] proposed IoT based door
access control using face recognition. The proposed system mainly consists of subsystems namely image
capture, face detection and recognition, email notification and automatic door access management. The
captured image from a raspberry pi camera will be sent to the authorized person through email for safety
purposes. In this paper the unauthorized users are not consider, to advance this we are proposed guest mode
in which guest could ask a permission and get response via email notification in order to access the room.

The authors in [5]had implemented authentication system where if any person came at the door it was
notified to the home owner via e-mail, then the owner could see the person standing at the door using
camera from remote location. This paper has a limitation when the camera difficult to detect the face. to
advance this we have introduce two factor password authentication system using keypad.

IJISRT22JUL1344 www.ijisrt.com 837

Volume 7, Issue 7, July – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
The Author in [6]have been proposed automatic attendance System Using Image Processing and deep
learning algorithm. Face detection and recognition section detect face from the image capture by the camera,
and the image of the face is cropped and stored. The system recognizes the images of student’s faces, which
have been registered manually with their names and ID in the record. The author only considers on marking
of attendance, but in this paper, we have been considering on both marking attendance and door open.

The author in[7] had been proposed intelligent door lock system with face recognition to control the
present barriers in conventional door locks. The system had been worked by detecting and recognizing the
human face with the OpenCV library on the Arduino platform. The proposed system has been elaborated to
stop robbery in extremely surveillance areas like home environment with less power consumption and more
reliable standalone security device for both intruder detection and also for door security.

With the flow of technology advancement, more efficient algorithm, methods and systems are proposed
and developed for robust service. In this project we have proposed computer vision-based authentication and
employee monitoring system. The system works both on gate controlling and attendance marking
simultaneously with an efficient face-recognition algorithm. Two factor password authentication system are
also used when the face recognition system difficult to detect the humans face during fog and night
conditions. From the survey, most researcher focus only on one condition either face recognition base gate
control, or attendance mark system on day time or password authentication system for an authorized user or
employees. In our case the guest mode also considered via email notification. Also fog and night conditions
are also considered which make the system multi-functional.

F. Methodology
In order to accomplish this project, we will follow a certain step and carry out different tasks. Different
literature has been revised relating to this project and detail analysis and implementation is used to reach
result and discussion. The general procedure in doing the proposed system is as follow in the figure below.

Pro Lite Dat Dat Dis

ble rat a a cus con
m ure coll ana sio clus
ide revi ecti lysi n ion
ntif ew on s and
icat con
ion clus
ion
Figure 1.1 Procedures we follow in the project

In this project, we will be using the LBPH (Local Binary Pattern Histogram) face detection algorithm.
This algorithm will give us more accurate results when we compare to other types of algorithms such as
Fisher Face, Eigen Faces algorithms. This LBPH Algorithm will take number of images as you wish in
different angles and check those all images at the time of face detection. In our case, we will be taking 30
images of a person with different angles and it will be stored in a database. We recommend to use a greater
number of images with having variations to have better result, but it has tradeoff with memory and processor
power. At first, it converts color images to gray scale images and then it converts into pixels for detecting
this will divides the image into various pieces then it stores the values of each pixel. Images will train and
store onto the database as encoded file.Pixels those grouped into lowand high will takes as 0 and 1.It will be
arranged into a 3 x 3 matrix format for recognizing the new images on screen compared to data base stored
images. From this we divide the phases into preprocessing, feature extraction and classification phases.

IJISRT22JUL1344 www.ijisrt.com 838

Volume 7, Issue 7, July – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
The first phases are pre-processing of the image, which refers to the gathering of the image data from
the camera module. But we do not need to save whole image in the dataset. We will only need a part of the
face from whole captured image. For this, we will have to detect the area of the face in the image. A short
code for face detection is developed. This code is also useful for other modules in the system. This detected
part of the image will be cropped and saved in the data folder. Also, care has to be taken to align the images
if they are shot from a different angle.

The second one is feature extraction phase, in which once we have got the images for training, we can
use the image algorithm to learn on this dataset. Depending on the size of data samples, the accuracy of the
classifier will vary. We applied the LBP method on image pixels by thres holding the 3 × 3 neighborhood of
each pixel with the center value and considering the result as a binary number. The last one is classification
phase which refers to testing of our face recognizer. We will do a real time video check to verify the
correctness of the trained model. For a new face input,the system will first extract its features and generate
binary patterns same as we did for the training images. After its completion, the input is given to the trained
recognizer to classify the image according to its training.

The designed system will follow the following steps to do the image processing.

Input image

Figure 1.2 flow of operation

In this chapter we have discussed overall introduction of the proposed system such as background of the
study, statement of the problem, objective scope of the project, literature review and methodology.In the
coming chapter we will discuss system requirement, architectural design and specifications.

IJISRT22JUL1344 www.ijisrt.com 839

Volume 7, Issue 7, July – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165

CHAPTER TWO

SYSTEM REQUIREMENTS, ARCHITECTURAL DESIGN AND SPECIFICATIONS

In this chapter we will discuss system requirement, architectural design and specifications. To perform
projects full functional requirement with an efficient performance, setting specification and requirements are
mandatory. Before purchasing a software program or hardware device, one can check system’s requirements
to make sure the product is compatible with his/her application.

A. System requirements
System requirement are the required specification that a device must have in order to perform its operation.
This includes the minimum and recommended requirements of both the hardware and the software. If some-
one use components with bellow recommended specification, there might be got unexpected result and the
system will be nonfunctional.The hardware and software requirements include operating power,
performance of operating system to operate some applications, storage or Memory unit, Input output port to
work with peripheral device, and minimum GPU for displaying result.
a) Functional requirements
In our system, we grouped the following functional requirements the system need to be meet:
 Face capturing by the raspberry pi camera and then it saved on the raspberry pi memory.
 Face detection - if there is a human near to the camera, then it will be detected.
 Face recognition- if the detected man is registered on the system, then the man will be recognized by
face recognition algorithm and it will allow the access.
 Send notification to Admin- if the guest wants to access the room, they must send a request to admin
via emails in order to get the password.
 Email verification to user- if there is low light intensity, fog and difficult to detect the face then the
system allows the user to access the system via two factor authentications.
 Automatically open the door when the authenticated person wants to access the system and the door
will be closed after users enter to the room.
 The LCD will display an instruction based on user’s request.
 Accepting input via keypad when the camera system is non-functional.
 Sensing the presence of human near to the camera system.

b) Non – functional requirements

We also grouped the following non-functional requirements to the system:
 Graphical user interface.
 Announcing employee when they are removed from the system.
 Safety and protective measures we aim to integrate including sensors for this purpose.

c) Hardware and Software requirements

In this section we mention hardware and software requirements of the system. we mentioned only the
name along their model of the components both hardware and software those we are going to use in
theproject. There detail spec will be mentioned in the Engineering design and spec section on chapter
three.
 Hardware requirement
 Raspberry PI3(Model B V1.2): Raspberry pi that contain camera slot, minimum of two USB port,
1GB RAM, multiple GPIO pin, micro-USB port and VGI port. Operating power is 5V with 2.5A.
 Pi Camera v2.1: a PI camera with a minimum lens size is 25x24x9, weight is 3g and its resolution is
8MP. The recommended resolution is using high resolution like 8MP and above.
 Servo motor:operating voltage will be 5V to 12V, inertial ratio is 3:1 or less, high performance
servo drives have auto-tuning, but most servo motor manufacturers recommend that the inertia ratio

IJISRT22JUL1344 www.ijisrt.com 840

Volume 7, Issue 7, July – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
be kept to 10:1 or less, although there are many applications that operate successfully at much higher
ratios.
 LCD display:LCD with operating voltage 4.7V to 5V, and contains Data pin for interfacing to
raspberry pi, and contrast pin for adjusting the level of brightness.
 PIR sensor: cover a minimum range 0.5 meter or above, with high accuracy
 4x4 Keypad matrix: support both alphabet (A,B,C,D), digits (0-9) and some special characters like
* and #.
 Bread board.
 5volt DC battery source.
 Wires and resistors.
 Desktop computer’s monitor: for displaying purpose.

 Software requirement
 Operating system: Debian Raspberry PI OS-32 bit and above.
 Python:Python environment with its library like OpenCV, pillow, NumPy and python GUI.
 PyCharm IDE: used as an editor for the codes.
 Windows OS | portable with Linux too: with minimum window 7, but the recommended is windows
10 and above.

Specifically, from the python environment, we use the following library

 OpenCV: OpenCV is an image and video processing library and is used for image and video analysis,
like facial detection.
 NumPy: NumPy is the fundamental package for scientific computing in Python which provides a
multidimensional array object. we used it to convert our images into some form of an array so that we can
store the model that has been trained.
 Haar Cascade: Haar-Cascade is basically a classifier which is used to detect the objects for which it has
been trained. The result is an XML file which stores the trained result. Simply the Haar-Cascade is
trained by superimposing the positive image over a set of negative images. We need haar-cascade frontal
face recognizer to classify faces into groups for recognition purpose.

B. Architectural Design and Specifications

In this section we will discuss the architectural design and specifications of the system.
a) Introduction
We used adata-flow graph (DFG)for representing/ modeling connection of physical components and
sequence functions used in the system. DFG is a collection of arcs and nodes in which the nodes are
either places where variables are assigned or used, and the arcs show the relationship between the
places where a variable is assigned and where the assigned value is subsequently used[8].

Below is a general system model, having an input, the system itself and the outputs.

System (Raspberry Pi)

Inputs Outputs
System (Raspberry Pi)

Inputs System (Raspberry Pi) Figure 3.1

General
Figure 2.1 general system description system
description.
Inputs System (Raspberry Pi)
Outputs
IJISRT22JUL1344 www.ijisrt.com 841
Inputs
Figure 2.1
Volume 7, Issue 7, July – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
A bit detail of the main system components is shown below.

input interface Output interface

input interface Output interface

input interface
Camera Raspberry PI ServoOutput
Motor interface

input interface Output interface

Camera ServoLCD
16X2 Motor
Keypad

Camera ServoLCD
16X2 Motor
Database(server)
PIR
Keypad

Camera ServoLCD
16X2 Motor
Figure 3.2 Database(server)
Keypad Figure 2.2 our system description
Our
b) Systems (Raspberry PI) detail architecture
system
In this
descriptio we try to express the relationship and sequence
section, 16X2 LCD
Database(server)
of components as well as functions
Keypad
(those
n -run on the system) interaction, along the data type they send during a communication by using
on the
DFG. Rectangular shape is used to represent hardware components such as servo motor, while
roughPIR
elliptical shape used to represent the software - functions(methods).
Database(server)
a. Detail about the data flow diagram
As usual
Figure 2.2we consider to start from the left side where an employee sends his facial data and to be
recognized by the system, or get an access to a protected resource after he get a confirmation by
Our
entering a onetimepassword (OTP) via a keypad. Once a user interfaced the input, his data such as
face image or passwords- those are grouped into alphanumeric data will accepted and processed
system
by going to the methods found in the main controller which again contains more than one
descriptio
method. Once the input are accepted the main.py python script will identify the method where
nthose
- on incoming
the input data should intend for. The methods inside the main controller are main.py
which is a super controller for the system, guest.py which manage the guest mode request by
rough
receiving a request from someone, send to admin email and send back admins review and so on,
otp.py which will manage the two-factor password authentication procedure. when there is no
threshold light for the PI camera this python script will process the authentication via a password
in which the system and the user communicate via an email address.
PIR
In the controller, the input data will go over a series of conditions, controls and loops for
recursive actions and after a while an output data will forward to the output section. To mention
this with
Figure 3.2 a typical example, if a registered employee face data much with the one stored in the
database,
Our a data/command will send to the motor interface which again trigger the motor to rotate
and open the door for a while. Even though the data flow diagram is depicted as an end to end
system
descriptiothe detail and real flow of the data is dynamic- at sometimes it may go to the beginning
system,
and do recursive processes. Such details are elaborated in the algorithm design section ion chapter
n - on the
three. Between a sequential two nodes, we tried to mention the data that is passed over. For data
those PIR
rough are known specifically face data between the camera and the controller and some gross data

IJISRT22JUL1344 www.ijisrt.com 842

Volume 7, Issue 7, July – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
those not mentioned explicitly due to the data passed particularly. Here one can think the data that
should go over is the destinations node method parameters , so we can think of it like this
one.Between the main controller and the hardware components, there is a layer (explicitly shown
with a dot rectangle in fig 2.2 above) called an interface which act as a link between the two ends.
Systems detail data flow diagram is attached in the appendix with A3 paper, please go and refer.

Data flow diagram, check at the appendix section.

Figure 2.3 signal flow graph

In this chapter we have discussed system requirement, architectural design and specifications. In the
coming chapter we will discuss the engineering design and detail specifications of the system. Also detail
algorithm design, system’s hardware connection and mechanical design will be covered.

IJISRT22JUL1344 www.ijisrt.com 843

Volume 7, Issue 7, July – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165

CHAPTER THREE

ENGINEERING DESIGN AND DETAIL SPECIFICATION

In this chapter we will mention detail engineering design and calculations along detail engineering
specification of components. And also, we will have mentioned the operating power of each component and
discussing how the algorithm works by indicating the flow chart of user registration, training and face
recognition system. Finally, the electrical diagram are drawn by using fritzing software.

A. Engineering design and detail specification

In this section, we will cover hardware detail engineering specifications and related calculations. The
calculation includes doing it via a simple logic like, if the raspberry pi sends an output of 3.3-5 volt, then it
is not enough to run a 12v motor, so we have to interface the Raspberry pi and the motor via a relay module
(one which may integrate with a motor drive or a separate one). It may also go to design a component from a
scratch by measuring all detail input outputs like doing a feedback control system for a particular purpose, it
needs doing basics calculation on transistors, amplifiers and so on.

a) Raspberry pi
The Raspberry Pi is a microprocessor like a single board computer and can host the operating system.
The Raspberry Pi runs Linux distribution OS, but it also provides a set of GPIO (general purpose
input/output) pins, allowing you to control electronic components for physical computing and explore
the Internet of Things (IoT).

Raspberry Pi OS is a free operating system based on Debian, optimized for the Raspberry Pi
hardware, and is the recommended operating system for normal use on a Raspberry Pi. In this project
we have used third generation quad core raspberry pi 3 model B V1.2, which have 40 GPIO pins,
Camera Serial Interface (CSI) camera port for connecting a Raspberry Pi camera, Display Serial
Interface (DSI)display port for connecting a Raspberry Pi touchscreen display, Micro SD port for
loading your operating system and storing data, and Upgraded switched Micro USB power source.

A Raspberry Pi must be powered with a compatible power supply. Raspberry pi 3 model B work
fine on 5volt with 2.5A. The board consists of two 5V pins, two 3V pins, and 9 ground pins (0V),
which are unconfigurable.

A powerful feature of the Raspberry Pi is the row of GPIO (general-purpose input/output) pins
along the extreme right edge of the board. GPIO pins do not have a specific function and can be
customized using the software. A GPIO pin that is set as an input will allow a signal to be received by
the Raspberry Pi that is sent by a device connected to this pin. A voltage between 1.8V and 3.3V will
be read by the Raspberry Pi as HIGH and if the voltage is lower than 1.8V will be read as LOW. On
the other hand, a GPIO pin set as an output pin sends the voltage signal as high (3.3V) or low (0V).
When this pin is set to HIGH, the voltage at the output is 3.3V and when set to LOW, the output
voltage is 0V. The raspberry pi contains both software and hardware pulse width modulation. Software
pulse width modulation is available in all pins, but hardware pulse width is available on specific GPIO
pins like GPIO12, GPIO13, GPIO18, and GPIO19.

For the raspberry pi and for the other devices, we summarized the detail spec in table following
the descriptions.

IJISRT22JUL1344 www.ijisrt.com 844

Volume 7, Issue 7, July – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165

Figure 3.1 raspberry pi and its pin configuration

b) Liquid Crystal Display (LCD)

A liquid crystal display (LCD) is a type of display technology that makes use of liquid crystals that
open or close when stimulated by an electric current. Out of various types of LCD, we choice 16×2
LCD which is cheap and easily to program. The 16x2 LCD can display alphabets and numbers
Consists of two rows and each row can print 16 characters. Each character is built by a 5×8-pixel box.
So, it will have (16×2=32) 32 characters in total. The operating voltage of this LCD is 4.7V to 5.3V
and current consumption is 1mA without backlight Alphanumeric.

A 16×2 LCD has two registers such as data register and command register. The RS (register
select) is mainly used to change from one register to another. When the register set is ‘0’, then it is
known as command register. Similarly, when the register set is ‘1’, then it is known as data register.
 The command register is used to store the instructions of command which are given to the display.
So that predefined tasks can be performed such as clearing the display, initializing, set the cursor
place, and display control.
 The data register is used to store the information which is to be exhibited on the LCD screen. Here,
the ASCII value of the character is the information which is to be exhibited on the screen of LCD.
Whenever we send the information to LCD, it transmits to the data register, and then the process will
be starting there. When register set =1, then the data register will be selected.

Figure 3.2 Liquid crystal display (16x2 LCD)

IJISRT22JUL1344 www.ijisrt.com 845

Volume 7, Issue 7, July – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
c) Relay module
A relay is an electrically switch, which is operated by electromagnet. It consists of a set of input
terminals for a single or multiple control signals, and a set of operating contact terminals. The
electromagnet is activated by a separate low-power signal from a micro controller. When activated, the
electromagnet pulls to either open or close an electrical circuit.

When the relay is de-energized, the sets of contacts that were closed, open and breaks the
connection and vice versa if the contacts were open. The operating voltage of relay in door application
is 12V or 24V and current is 500mA or 250mA respectively.

Figure 3.3 One channel relay modules

d) Servo motor
Servo motor is a combination of DC motor, position control system and gears. The motor has shaft,
the rotation or the angular position of which is controlled by the signal that appears on the signal line
of the servo motor. The servo motor has three wires connected to it. Two of three wires of the servo
motor are the power pins and the third pin is the signal pin to which the signal is connected in such a
way that the pattern of the signal controls the angular position of the shaft of the Servo Motor. we are
surveying weights of many international doors, so the weight range is between 5kg to 40 kg.Stall
torque is the torque produced by a mechanical device whose output rotational speed is zero. It may
also mean the torque load that causes the output rotational speed of a device to become zero.A
9.4kg/cm Servo motor should be able to lift 9.4kg if the load is suspended 1cm away from the motors
shaft, the greater the distance the lesser the weight carrying capacity.For this project demonstration we
have used SG90 Servo Motor. It is one of the popular, cheapest and 180-degree servo. Its operating
voltage is 4.8V.

The rotation angle of the servo motor is controlled by applying a Pulse Width Modulation
(PWM) signal to it. By varying the width of the PWM signal, we can change the rotation angle and
direction of the motor. In servo, we have a control system which takes the PWM signal from Signal
pin. It decodes the signal and gets the duty ratio from it. After that, it compares the ratio to the
predefined positions values. If there is a difference in the values, it adjusts the position of the servo
accordingly. The frequency of PWMsignal for SG90 is 50Hz.

PWM is a type of signal which is obtained from Raspberry Pi Pico board. The output signal is a
square waveform which at a particular instance is either high or low. If we are using a 3.3V power
supply then the PWM signal would be either high which is 3.3V or low which is 0V. The ‘on time is
the duration till which the signal stays high and the ‘off time’ is the duration till which it stays low.

IJISRT22JUL1344 www.ijisrt.com 846

Volume 7, Issue 7, July – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
𝑜𝑛 𝑡𝑖𝑚𝑒 𝑜𝑓 𝑠𝑖𝑔𝑛𝑎𝑙
Duty-cycle =𝑡𝑜𝑡𝑎𝑙 𝑡𝑖𝑚𝑒 𝑝𝑒𝑟𝑖𝑜𝑑 𝑜𝑓 𝑠𝑖𝑔𝑛𝑎𝑙

Figure 3.4 a typical SG90 servo motor

Inertia ratio is the most important factor for sizing servo motor. It can be used as a measure of how
well the motor is able to control the acceleration and deceleration of the load. Inertia ratio is defined as the
ratio between the inertia of payload to the inertia of the motor.
𝐽𝐿
Inertia ratio = 𝐽𝑀, where JL is the inertia of load, and JM is the inertia of the motor

Manufacturers typically provide the inertia value of the motor, so we need calculate the inertia of the
load. To determine the inertia of a screw-driven load, the effect of the screw’s lead must be considered.

Inertia of load driven by screw

where, m is mass of the load and Ls is load of screw

An inertia ratio that is too low means the motor is likely oversized, leading to higher than necessary
cost and energy usage. An inertia ratio that is too high means the motor will have a difficult time controlling
the load, which results in resonance and causes the system to overshoot its target parameter (position,
velocity or torque).

e) PI cameramodule
The Pi camera is a portable light weight camera that supports Raspberry Pi. It communicates with Pi
using the mobile industry processor interface (MIPI) camera serial interface protocol. It is normally
used for capturing videos in the system. Apart from these modules Pi can also use normal USB
webcams that are used along with computer. We are using 8MP resolution camera. It’s capable
of 3280x 2464-pixel static images, and also supports 1080p30, 720p60 and 640x480p90 video. But we

IJISRT22JUL1344 www.ijisrt.com 847

Volume 7, Issue 7, July – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
recommend to use camera above 8MB resolution for sake of image quality which has effect on later
processes.

Figure 3.5 camera module and its pin configuration

f) PIR sensor
A passive infrared sensor (PIR sensor) is an electronic sensor that measures infrared (IR) light
radiating from objects in its field of view. PIR sensor detects a human being moving around within
approximately 10m from the sensor. This is an average value, as the actual detection range is
between 5m and 12m. The operating voltage of PIR sensor For a D.C. works from 3.3V to 5V.

The PIR sensor itself has two slots in it, each slot is made of a special material that is sensitive
toIR. When the sensor is idle, both slots detect the same amount of IR, the ambient amount radiated
from the room or walls or outdoors. When a warm body like a human or animal passes by, it first
intercepts one half of the PIR sensor, which causes a positive differential change between the two
halves. When the warm body leaves the sensing area, the reverse happens, whereby the sensor
generates a negative differential change. These change pulses are what is detected.

Figure 3.6 a. Heat source movement b. output signal

g) 4 x 4 Alpha-Numeric keypad
The 4 x4 keypad module consists of 16 keys, these Keys are organized in a matrix of rows and
columns. Normally there is no connection between rows and columns. When we will press a key, then
a row and a column make contact. When we will Pressing a button shorts one of the row lines to one
of the column lines, allowing current to flow between them. For example, when key ‘Button 1’ is
pressed, column 1 and row 1 are shorted, then the character is sent to the required output.

IJISRT22JUL1344 www.ijisrt.com 848

Volume 7, Issue 7, July – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165

Figure 3.74x4 alphanumeric keypad and its internal structure

h) Screen Display-Monitor
A display device is need to show and display the designed GUI for the admins in a central room. A
simple desktop monitor or any screen display devices those can support HDMI cable can be used. In
our demo, we are using desktop’s monitor as a display.To specify, a monitor having a moderate
resolution level and screen size of minimum 15inch is enough.

i) Server Computer
We were doing and thinking on how we can make the project more relevant by making it expandable
and following we were searching on how we can manage and do all logical operations performed on
the existing -demo level inside a remote, powerful, more efficient server computer. This will solve the
processing capability of the raspberry pi for hundred thousand of users and we are approaching to
implement a remote server. For this scalable scenario, a separate remote server that will manage and
do tasks those were done in the raspberry pi such as storing data including faces and metadata
including user name, user id, email and others on remote server. Both admins and the raspberry pi will
access the remote server. and each data modification will synchronized. Since both are not in one
room-the usual one, there need to have a network where both ends will communicate over. The
protocol over the network will depend on the organizations, it is better if the network which is going to
created is compatible and inclusive within the existing resources. Below diagram shows the way in
which the local raspberry pi and remote servers are communicated.

Local machine Remote server

(Raspberry PI) Network (SSH or HTTP,) computer

Figure 1.8. communication

Figure 3.8 communication between the local between the local device
device raspberry pi and the remote server via network
raspberry pi and the remote
We can summaries the device spec as shown table
server via in the next page,intable one for the design level and
a network.Network
table two for the implementation part: (SSH or HTTP,..)

IJISRT22JUL1344 www.ijisrt.com 849

Volume 7, Issue 7, July – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
Device name Specification
Raspberry pi - Quad Core 1.2GHz Broadcom BCM2837 64bit CPU
- Support: Ethernet cable,
- Full-size HDMI CSI (15-pin Camera Serial Interface) camera port for
connecting a camera
- RAM: 1GB and more, processor
- Micro USB-2 power port (up to 2.5A)
- Micro SD port 4GB and above
- Supply voltage 5V
- 40-pin extended GPIO
- DSI (Display Serial Interface) display port for connecting a touchscreen display
- 4 USB 2-port
- Video: Supports 1080p @30fps, 720p @60fps and 640x480p 60/90 Recording
LCD - Operating Voltage is 4.7V to 5.3V
- Current consumption is 1mA without backlight
- Can work on both 8-bit and 4-bit mode
- display alphabets and numbers
Camera - Resolution: 3280 * 1944 /8MP and above color camera module
- Supports: 1080p, 720p and 480p
Relay - supply 5v and above
- maximum contact voltage – 250VAC or 30VDC
- maximum current – 10A
- Relay is activated 70mA
PIR sensor - Range 2-4 meter
- Good accuracy
- Level of output: high 3.3V, low 0V
- Operating voltage 5V
Servo motor - Stall torque min: (5V) 21Kg/cm
- Stall torque min: (6.8V) 25Kg/cm
- Operating speed: 0.15 s/60º (5 V), 0.16 s/60º (6.8 V)
keypad - 4x4 alphanumeric
- Maximum Voltage rating 24V
- Maximum Current rating 30mA
- Numbers of Pin 8
Table 3.1 hardware specification fort the design-level

Device name Specification

Raspberry pi - Similar with the above.
LCD - Similar with the above specification
- 16x2
Camera - Resolution: 2592 * 1944 / 4MP
PIR sensor - Similar with the above,
Servo motor - Operating voltage 4.8V
- Weight 9gram
- Torque 1.8kg-cm

keypad - Similar with the above,

Table 3.2: Hardware specification (for the implementation-prototype).

IJISRT22JUL1344 www.ijisrt.com 850

Volume 7, Issue 7, July – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
B. Algorithm Design
Algorithm is a set of procedures inorder to do some tasks. Algorithms are never written to support a
particular programming code.We used a flow chart to describe the algorithm of the system where the set of
procedures the system will go through to do the expected tasks. The end task the algorithm aims to destined
is either recognizing an employee via his face and by going several steps, after recognition done, the door
will open and the user/employee is authenticated and the algorithm ends at this point. In another perspective,
if we are considering a registration and training phase of the model for new user/ employee existed, the
algorithm aimed destination after going several steps, model training is done and a trainer file will generate
which is face’s encoded data file and the algorithm ends at this point. Since it is easy to recognize and
understand the flow and relations, we used flowchart over pseudocode.

Flow chart for Registration with training and for the recognition process of the system is give below.
The drawing is done on flowchart-draw.io, online1.

Figure 3.9 flow chart for registration and training

1
Flowchart-draw.io link: https://app.diagrams.net/

IJISRT22JUL1344 www.ijisrt.com 851

Volume 7, Issue 7, July – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165

Figure 3.10 flow chart for face-recognition

C. System’s hardware/ circuit configuration(implementation)

Below we try to draw the possible hardware component connection and their procedure for the designed
system as shown below by using a fritzing drawing application.

IJISRT22JUL1344 www.ijisrt.com 852

Volume 7, Issue 7, July – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165

Figure 3.11 System’s circuit connection (drawn in fritzing)

Here, the left bottom green device is the raspberry pi, the lines in between devices are numbers of wires
those are going to draw between for data/signal exchanges. The left top sensor is a passive infrared ray
sensor which used to detect the presence of a person near the camera. In the right side, a 4x4 keypad matrix,
a relay module, DC (5volt) battery, LCD along a potentiometer for adjusting its brightness. In the between a
bread board is placed.

D. Mechanical Design
Below we try to draw the door system where the system is going to operate on. We draw it by using
draw.io software.

Figure 3.12 mechanical design of the door system

IJISRT22JUL1344 www.ijisrt.com 853

Volume 7, Issue 7, July – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
A more clear and real door which shows the closing and opening mechanism is show below.

Figure 3.13 the mechanical design for controlling door movement

a) Working operation
Initially before the motor gets command form the microcontroller, the door is closed and the first link
that is attached to the servo motor inclined at an angle of 45 0 as shown in diagrammatic picture bellow.

Figure 3.14 initial position of the door (door is closed)

When the servo motor rotates in the clockwise direction, the first link will also rotate in the same
direction as the motor, the rotation of the first link pushed the second link and the second link pulls the
sliding bar that holds the door then the door will starts sliding to the right

IJISRT22JUL1344 www.ijisrt.com 854

Volume 7, Issue 7, July – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165

Figure 3.15 the sliding door start to open

Final, the motor continues rotating and the link attached to the motor riches at angle of 135 0 the door
will fully opened.

Fully opend

13
50

Figure 3.16 the door reaches final position (opened)

Once the link attached to the motor reaches 135 0 from the horizontal, the servo motor will finish its
rotation in the clockwise direction and it is ready for the anticlockwise direction to close the door. For
closing of the door, a similar operation will be followed in the anticlockwise rotation of motor.

Figur 3.17 initial to final movement of the firist link

IJISRT22JUL1344 www.ijisrt.com 855

Volume 7, Issue 7, July – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
In this chapter we have discussed the system’s engineering design, detail specifications, algorithm
design and mechanical design. In the coming chapter we will discuss implementation, testing and
discussions. In the implementation section we cover component connection of the circuitry and detail code
explanations.

IJISRT22JUL1344 www.ijisrt.com 856

Volume 7, Issue 7, July – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165

CHAPTER FOUR

IMPLEMENTATION, TESTING AND DISCUSSION

In this chapter we will elaborate the topics about our implementation of the designed system, the detail
procedure for the operation to be followed and the detail behavior on which the data flow diagrams and the
algorithms designed in the previous chapter will be entertained with a real input data and each phase of the
algorithm and conditions will be testified in a logical coherency. At each stage, we will test and document
the results and along each result and procedures, each phase will be discussed with a detail illustration.

A. Implementation
The entire system is made up of raspberry pi, LCD display, PIR sensor, Pi camera module, Relay module,
servo moto, keypad, electrical circuit board and the gate system. The raspberry pi a single board circuit
(SBC) microprocessor that control all activities inside the system.

The system contains two phases. The first phase is so called registration and training phase and the
second phase is called recognition phase. during the first phase the system’s admin registers the user’s
information user’s such as Identification number, full name, gender, and email address. After filling the
form, the admin clicks the camera activating button (take image button), then the PI camera capture 30
images and store it on image directory. After registering the user, the system admin trains the model over the
images found in the image directory over the whole registered users face. The training phase will take the
image on the image directory as an input and by having the haar-classifier, it will extract a feature for each
face images and once it correlates and associate, an encoded face data will be stored and save in a trainerfile.
This generated .yml file will be used in the second phase for the recognition stage to compare the images
extracted data. For every new user registration, a training need to have performed, else the system will have
no the new user’s facial data and will probably know as “Unknown!”.

When we come to face recognition phase, the PIR camera sense the presence of human and if it is
present, the PI camera is trigger and start detect humans face. If the detected face is matched with the face
data stored in the trainer file – which is considered as it saved on a database, then the name of the recognized
person is attached with his/her face. during this time, an LCD display’s method will be called and passed a
string - ‘the door is open, welcome Dear user!!’ at that time the system main controller wills send a
command to the raspberry PI GPIO pin where the motor drive is listening a data, and will forward a
command to open rotate the motor and the door will opened for a while. Again, attendance is also marked by
including, full name and id of attended person with attended date and time. If the detected person is
unknown then, the person must ask a guest mode and send request to admin via email, and then the admin
give a permission to access the door, unless the access is denied. The trainerfile is used as a reference
during face recognition system.

For the sake of resources, time and budget constraint, we are going to implement only the local system
which consist of the whole electrical components presented in the electrical connection as given before in
the prototype’s circuit connectivity.

The raspberry pi which is the central system, also known as the brain of the system. It is used to store
data, control the I/O pins, monitoring users. The PIR sensor is used to activate a camera system by detecting
the presence of humans within a specified range. Since raspberry pi’s GPIO pins release a maximum of 3.3
volt only, and insufficient to run an external servo motor and for the sake of raspberry pi’s safety due to
back electromagnetic force, a relay module is usedto switch the servo motor during closing and opening the
gate. The alphanumeric keypad is an input of the system that is used to enter a string in to the system. Such
strings can be user’s password and username, admins password and username, or a onetime password given
by the admin. The LCD display is an output of the system that used to display some instructions to the user.

IJISRT22JUL1344 www.ijisrt.com 857

Volume 7, Issue 7, July – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
The LCD is interfaced with the raspberry pi via implicit pull up/down resistors those are built inside the
PI and only an external potentiometer is used to adjust the brightness of the display. As depicted in the
algorithm for the recognition scenario, the very beginning task/process goes is the PIR sensor detection for
the presence of a human around the gate. The connection for the IR is, it gets a 5volt input and the data pin
is connected to raspberry PI’s GPIO pin number 17. The code is written in a way to send a data and activate
the camera then the whole system to detect, recognize and do the rest functionalities. Once the system is
started, it will grape a frame from the employee face and check face’s extracted feature with the one stored
in the database, after making a series of procedures, the system will validate either the employee has to
entered or not. The grippedimage will be passed over the haar-classifier and a feature will be extracted. If a
match found, the PI will send a command data to the motor interface and the motor; will opened for a while
and allow the employee to entered. If the face take is out of the database facial data, the user will get a guest
mode request. This is nothing but, for a user who has no registered yet on the system gets a trial to get access
after having a permission from am respective admin of the system. The code is written to send a request
from the employee to the admin, sending the response from admin to the employee via a simple mail
transfer protocol (SMTP)- email services. Here, we are only reliable on the security which google provide in
the communication protocol which is secured shell transmission (SSL) and for the case those need a greater
security on such communication, another encryption technique can be deployed on it.

When the PIR’s data shows a presence of a man in front of the camera, but no face detected, the case
can be for two reasons, one it may not have enough brightness level which the camera needs to work
properly. The second and most rare scenario is when the user faced the camera in inappropriate position. We
only consider and integrate the first- at night time and in a summer season, a fogy whether cause the
brightness to dimension. During this case, the system will trigger the users to gain access to the resource
once they’re authenticated via a two-factor authentication. As we brief the registration phases
implementation, a new user will give his email, password, username during his registration. So, the system
will prompt the user to follow his email and once he requests this OTP means of authentication, the system
will generate a one-time password and send to his email and if here entered he correct username and
password related to his email address, he will get access to enter the door once the entered pass code is
similar to that generate by the system within a short period of times. The generated OTP will be stored in a
database and once it will used, the system clears it. But in our implementation, we only store the OTP in a
script file that is on the run on a variable and the best and scalable way id the one which stored in a database.

a) Component connection through GPIO pins

When we come to the implementation detail, the PIR sensor, the PI camera, the 4x4 matrix keypad,
LCD, relay module, and the desktop are connected to the central unit called raspberry pi. The
raspberry pi camera’s 15 pins are connected to the raspberry pi camera slot port.

The PIR’s ground pin connected to the GND pin of raspberry pi, the VCC pin of PIR is connected
the VCC of raspberry pi and the third pin is data pin, which is powered by the 3.3V raspberry pi
GPIO5 and send the presence of human to raspberry pi by using this data pin.

The 4x4 matrix keypad’s row R1, R2, R3, R4 are connected to GPIO26, GPIO19, GPIO13,
GPIO6 pins of PI and the column C1, C2, C3, C4 are connected to GPIO12, GPIO16, GPIO20,
GPIO21 pins of raspberry pi respectively. The users are communicating or send a steam of character
to the system through these pins.

For the LCD, first, connect pin GND and K of the LCD to GND and VDD and A pin to 5V
supply. Then connect a 10KΩ Potentiometer to V0 pin of the LCD, which is the contrast adjustment
pin. The two control pins of the LCD i.e., RS and E are connected to GPIO7 and GPIO 8 respectively.
Now, the data pins of the LCD, since we are configuring the LCD in 4-bit mode, we need only 4 data
pins (D4 to D7). D4 of LCD is connected to GPIO25, D5 to GPIO24, D6 to GPIO23 and D7 to
GPIO18.

IJISRT22JUL1344 www.ijisrt.com 858

Volume 7, Issue 7, July – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
Connecting electric motor directly to the raspberry pi may cause a damage due to high current in
the motor, we need to insert relay module between the raspberry pi and the servomotor. The output of
the raspberry pi GPIO 22 is connected to the signal pin (IN) of the relay, so the microcontroller
controls the relay via this GPIO pin. the Ground pin of relay is connected to the ground pin of the
raspberry pi and the 5V VCC pin is connected to 5V VCC of raspberry pi. When we come to the other
side of the relay module (output side), the normal open pin is connected to VCC of power supply (5V
or above) and the common of relay is connected to the VCC of servo motor.

The ground pin of the servomotor is connected to ground pin of power supply. On the other hand,
the signal pin of a servo motor is directly connected to the PWM port of the raspberry GPIO17. A
servo motor is controllable through PWM. Specifically, its arm position depends on the width of the
pulse applied to it. Once a match face is found, the Raspberry PI’s GPIO will send a command to the
motor drive (relay) and the motor dive (relay) will switch the relay to run the servo motor.

Below is summary of component connection

Table 4.1 component pin connection

b) GUI and python Scripts uses

Python has a lot of GUI frameworks. Tkinter is a graphical user interface (GUI) library for Python,
used to make buttons, desktop apps, image and text area features with an abstractive feature. The
system has an interactive User interface as mentioned in a screenshot below by using python-tkinter.
The detail on how the GUI is accomplished will explained in the code’s description in the coming
section.

Following, we are going to explicitly mention and explain the working procedure and the
implementation by having a reference with codes we wrote. Here we will take only a snippet of code
for illustration purpose. The whole project code is protected for copy right and members interest to

IJISRT22JUL1344 www.ijisrt.com 859

Volume 7, Issue 7, July – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
expand and do the project in the coming time with the scalable version, hence any interested party can
contact us and we can work together on it under some conditions.

To tell the whole with a bird view, the project codes are organized in a separate python script to
manage and handle easily by calling a method via their class’s instantiate- object. We have the
following python scripts:
 Headshot.py  Otp.py
 Recognition.py  Guest.py
 Train_model.py  Admin.py
 Database.py  Main.py

c) Detail code explanation (implementation)

Here for all the python scripts, we import various python’s libraries those are related to computer
vision and some others those help to manipulates data. These includes python’s tool – OpenCV,for
image processing and performing computer vision tasks.We install the package into the raspberry pi
python environment, and once installed we used it in a python script by importing it as its aliasing
name called “cv2”.
Import cv2
Other libraries like OS- which used to manipulate file management and other related tasks, time,
SQLite, tkinter, and raspberry pi’s GPIO are imported in their respective scripts. The detail is attached in
each script below.

TheOS module in Python provides functions for interacting with the operating system.This module
provides a portable way of using operating system-dependent functionality like manipulating
directories/paths.

TKinterprovides a powerful object-oriented interface to the Tk GUI toolkit.

Time module is used to manage time related tasks.

GPIO pins are imported to declare the pins and related functions in the python script.

SQLiteprovides a lightweight disk-based database that doesn’t require a separate server processand
allows accessing the database using a nonstandard variant of the SQL query language.

Below we discussed each python scripts as follow:

a. Headshot.py
This python script’s main task is to initiate the camera and store/grape images from an employee’s
face by the help of the OpenCV package.Inside, there is one class named Headshot, and two
functions named __init__ which used to consider the folders and the scripts as a python package for
later usage by importing it and start_headshot, which contains all the detail to take image from the
face. From this script, a snippets code which loop over the frames of person’s face and putting in an
array of data which will be stored later in a .jpg file is given below:

IJISRT22JUL1344 www.ijisrt.com 860

Volume 7, Issue 7, July – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165

And a code to save a gripped image into image directory:

The tkinter window need to be closed one a script finished the task as:

cam.release() #destroying the videocam.

cv2.destroyAllWindows() #destroying the started tkinter windows.

b. Train_model.py
In this script, additional packages such as NumPy which is a Python library used for working with
arrays; threading, pandas which used for data manipulation and analysis in particular operations for
manipulating numerical tables and time series;PIL python image library which adds support for
opening, manipulating, and saving many different images file formats.

The script has a class named Trian and methods are code to perform training model by looking
over the whole image directory and finally generating a trainer file. With the help of the OS module,
images will be traversed from their respective directory. A snipe code which trains the images on the
haar-cascaded classifier and saving the result into .yml is given below:

IJISRT22JUL1344 www.ijisrt.com 861

Volume 7, Issue 7, July – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
c. Register.py
In this script, the main task performed is registering new employee by receiving his information like
name, gender, email and etc. by using the GUI which relay on the TKinter library. The hardcode for
both the logic and the for the GUI is also goes here. The script has a code which validate the users
input like passwords and it also ensure the passwords are confirmed before saving with a single trial.

d. facial_reg.py
This is all about recognition- identifying either an employee is a known person or not. By taking a
face – which is extended from the startHeadshote class. Once the match is done, this method will
send commands to the main controller to do following tasks, opening a door or denying the request.
Since the comparison needs the database, we import the databse.py script- package as:

from Database import Database

Below we attached a snipe of code which loops on the employee face data and by extracting features
from it, comparing along the recognizer from the database encoded file-trainer, and finally
displaying the image with his/her name on the tk windows, which gives an access to the employee.
Else it displays, no match found on the tk windows along the face.

e. mailing.py
This script is all about managing the two-factor authentication which will communicate between the
system itself and the user via an email protocol – SMTP.For such purpose, we import the necessary
python’s packages called smtplib which defines an SMTP client session object that can be used to
send mail to any Internet machine with an SMTP and email which is used to manage email
messages.

Again, since we need to interact with the data stored in the database, we import the database
class here too. First the user will be screened via his username and email password, once he passed
this stage of verification, user send request for an OTP means of authentication, the system generate
an o and will send to his /her email and the user will enter this token via the keypad and he will get
access to the door system. The one-time password is expected to returned within 30 seconds, else it
will be expired and the session will be terminated.

A snipe code for a method named verify whichverify either the entered password is similar with
the one generated with in the specified time for the OTP to be valid is attached below.

IJISRT22JUL1344 www.ijisrt.com 862

Volume 7, Issue 7, July – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165

f. keypad.py
This script manipulates the users input like password and username and one-time password for
authentication purpose. It has imported the Raspberry Pi’s GPIO pins to listen and to write data to
the pins via the keypad matrixes. The code below ensuresthe configuration of the input pins to use
the internal pull-down resistors. C1,C2,C3, and C4 are variables those store pin numbers.

g. lcd.py
This script will manage all task those are related to display some alphanumeric data on the 16x2
LCD. As in the keypad.py file, PI’s GPIO pins are imported for the same reasons. Once you initialize
the LCD pins and related configurations in the raspberry pi, methods such as LCD strings are used to
display a string on the LCD display. The snipe code is attached below:

IJISRT22JUL1344 www.ijisrt.com 863

Volume 7, Issue 7, July – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
h. main.py
This script consists of the main part where the starter of the whole system is coded. It mainly
contains the admins login page. It asks admin to give his password and username to get access to the
system with admin level privileges. The GUI also embedded here in this script too for the login
windows. The passwords and the username are considered to be retrieved from a database, but for
our case, we just check from a dictionary-python’s data type, which work with value-pair
combinations of username and passwords. A snipe of code which checks the username and the
password given by the admin with the one stored in the dictionary is attached below. If they are
similar, the login page will call the main systems window- from the masterclass.

i. masterwindow.py
This script file is the central and most basic part of the project where most of the admin’s GUI is
coded, main tasks like registering new employee, and then training the model were called from a
separate window from this script.

Here, the script also code in order to show attendance in the front windows, along aGUI button those
are linked to a separate window to delete and to see the detail including employee lists. Below is a
code to send a notification message as the account of the user is removed from the system by the
admins for some reason. The code invoked the mailing class to send the notification via an email.

j. databse.py
The database python script is particularly coded to manage database related tasks only. The script
performs tasks including storing employee’s primitive data such as name, password, email etc. It also
used to see and check a previously stored data by retrieving from the database via the SQLite-
supported query language. Below code create an employee table, if not exist, adding a new employee
table.

IJISRT22JUL1344 www.ijisrt.com 864

Volume 7, Issue 7, July – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165

Below a snipe of code is given which will retrieve the data of the employee, particularly only his/her
name where his id match with the one passed with the method’s arguments.

Having all these implementation details, the circuit we are connecting and wiring in the implementation
is shown in the following attached pictures.

Implementations circuit connection, for clarity drawn on fritzing app.

a) sample implementetion circuit connectivity

IJISRT22JUL1344 www.ijisrt.com 865

Volume 7, Issue 7, July – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165

b) sample hardware implementation

Figure 4.1. Connection of Keypad, LCD, and PI camera with the Raspberry Pi. [a) drawn on fritzing, b)
image taken by our phone camera.]

k. Graphical User interface screenshots

In this section, we attached the results we got in the graphical user interface when the project
runs.From the very beginning the admin page seems as shown in the screenshot below:

Figure 4.2. admin login page

Below, the screenshot shows the home windows where New Registration, total list, remove, generate
attendance, generate report are linked via a button. Each button is linked with a separate window for detail
show.

IJISRT22JUL1344 www.ijisrt.com 866

Volume 7, Issue 7, July – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165

Figure 4.3. Home windows

B. Testing and Discussion

In this section, we will mention and present the results we got by testing the designed system via the
described implementations with real input data.

a) User registration
The admin register users in to the system by filling the form as shown in the figure below. After filling
the user’s name, email, Id, the admin clicks “take photo button”, in order to capture the users profile.
in this case 30 image are captured and saved on the directory Dataset with a file name the employee
index.

Figure 4.4 new employee registration window

IJISRT22JUL1344 www.ijisrt.com 867

Volume 7, Issue 7, July – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
b) Training the model

Figure 4.5 training the model(change the image pease)

c) Face recognition

a).

b).
Figure 4.6 Person face recognized, single person(a) and two persons (b)

IJISRT22JUL1344 www.ijisrt.com 868

Volume 7, Issue 7, July – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
d) Email verification (OTP) during fog weather condition

Figure 4.7 Email verification (OTP) during fog weather condition

e) Guest mode email notification

Figure 4.8 Guest mode email notification

f) The case when admin remove the user from the system

IJISRT22JUL1344 www.ijisrt.com 869

Volume 7, Issue 7, July – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165

Figure 4.9 admin remove the user

In this chapter we have covered implementation, testing and discussion of the system. In the last
chapter, we will cover conclusion and future works .In the future work we included a possible gaps and
limitations where the project can be improved by addressing those gaps.

IJISRT22JUL1344 www.ijisrt.com 870

Volume 7, Issue 7, July – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165

CHAPTER FIVE

CONCLUSION AND FUTURE WORKS

A. Conclusion
In this paper we have implemented computer vision-based authentication and employee monitoring
system. the authentication is based on face detection and recognition by taking the face images from the
streaming video with raspberry pi camera. The user will access the system when they are recognized. The
project has user mode and guest mode authentications. The guestneeds to ask a request from the admin with
an email notification to get access grant. For employee monitoring, the system automatically generates an
attendance when the user accesses the room or offices. This will minimize the drawback of manual
attendance system.

We have implemented two factor password authentication system with email notifications when there is
a fog or smog weather conditions are occurred. In the methodology the training image are stored on the
database. we have used local binary pattern histogram LBPH algorithm for face recognition system. the
algorithm is work based on a local binary operator, designed to recognize both the side and front face of a
human. Finally, for the testing case, when the motion is detected, the camera is initialized. The faces are
compared with the previously trained image. If the face is Match, then the system will be recognized, the
door is opened and the attendance is generated simultaneously. We had come across designing, developing
and testing a face recognition system which is computer vision-based authentication and employee
monitoring system.

B. Future work
For the future work, we hope the gaps we mentioned above and some limitations we got can be integrated
and done in the future including the remote server. To recap, the remote server isa scenario which isan
improved version over the project we are implemented which uses a separate remote server that will manage
and do tasks those were done in the raspberry pi for efficiency. Storing data such as faces and metadata
including user name, user id, email and others on remote server. Both admins and the raspberry pi will
access the remote server and each data modification will synchronized. The raspberry pi still will do the
main processing task such as the image processing part in detecting human face.

Level of light intensity to detect an image via the camera is very challenging task, it may be considered
as the hard part of computer vision, even we tried to calibrate and fix the issue with some techniques, still it
needs another level of improvement to detect and grippe an image from face in any light intensity
environment. In the future, there may be a special camera for face recognition, which can improve the image
quality and solve the problems of image filtering, image reconstruction, denoising etc. We can also use 3D
technology to supplement 2D images to solve some problems such as rotation and occlusion.

IJISRT22JUL1344 www.ijisrt.com 871

Volume 7, Issue 7, July – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165

REFERENCES
[1.] "computer vision,," [Online]. Available: https://www.ibm.com/topics/computer-vision.
[2.] "Face recognition,," [Online]. Available: https://customers.pyimagesearch.com/lesson-sample-what-is-
face-recognition/.
[3.] "openCV," [Online]. Available: https://opencv.org/about/.
[4.] H. Prasad, International Research Journal of Engineering and Technology , p. 4, May 2019.
[5.] Pavan Reddy Punnam, Dr.Munaswamy Pidugu, "Design of an Embedded Surveillance System,"
International Journal of Emerging Trends & Technology in Computer Science (IJETTCS), vol. 7, p. 6,
May - June 2018.
[6.] Sharvani Yedulapuraml, Rajeshwarrao Arabelli, Kommabatla Mahender, Chintoju Sidhardha,
"Automatic Door Lock System by Face Recognition," IOP Conf. Series: Materials Science and
Engineering 981 (2020) 032036, p. 8, 2021.
[7.] K. Tarun Reddy, K. Murali, M. Samba Murthy, G. Pavan, "Intelligent Door Lock System with Face
Recognition," International Journal for Research in Applied Science and Engineering Technology
IJRASET, vol. 8, no. V May 2020- Available at www.ijraset.com, p. 10, 2020.
[8.] "Data Flow Graph," [Online]. Available: https://www.sciencedirect.com/topics/computer-science/data-
flow-graph.

IJISRT22JUL1344 www.ijisrt.com 872

Volume 7, Issue 7, July – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
APPENDIX

The code for main.py python script [full project code is available from group members, upon request]:
### code start here ###
import tkinter as tk
from PIL import Image,ImageTk
from masterWin import MasterClass
class Main:
def login(self):
print("testing")
window1 = tk.Tk()
window1.title("AASTU -->Login window")
# get height and wodth to make the frame full screen
width = window1.winfo_screenwidth()
height = window1.winfo_screenheight()
window1.geometry("%dx%d" % (width, height))
# bg = tk.PhotoImage(file="bg.png")
img = (Image.open("bg.png"))
resized_image = img.resize((width, height), Image.ANTIALIAS)
bg = ImageTk.PhotoImage(resized_image)
cavana = tk.Canvas(window1, width=width, height=height)
cavana.pack(fill='both', expand=True)
cavana.create_image(0, 0, image=bg, anchor='nw')
# cavana.create_text(200, 200, text="joojo")
astuIcon = tk.PhotoImage(file=r"aastuIcon.png", height=500)
# labelIcon = tk.Label(window1, image=astuIcon).pack()
# labelTitle = tk.Label(window1, text="CV Based Authentication and Employee monitoring
System", fg='blue',
# font=('times', 30, 'bold'))
# labelTitle.pack()
# cavana.create_window(160+width/2, 100, window=labelTitle)
frame = tk.LabelFrame(window1, fg='red', text="For only the admins", padx=20, pady=20,
font=('times', 15, 'bold'))
frame.pack(padx=20, pady=20)

IJISRT22JUL1344 www.ijisrt.com 873

Volume 7, Issue 7, July – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
cavana.create_window(width/1.5, height/1.3, window=frame)
label1 = tk.Label(frame, text="UserName", font=('times', 10, 'bold')).grid(row=0,
column=0)
cavana.create_window(100, 100, window=label1)
self.uname = tk.Entry(frame, text="UserName", font=('times', 10, 'bold'))
self.uname.grid(row=0, column=1)
# cavana.create_window(100, 100, window=self.uname)
label2 = tk.Label(frame, text="password", font=('times', 10, 'bold')).grid(row=1, column=0)
cavana.create_window(100, 100, window=label2)
self.passkey = tk.Entry(frame, show="*", font=('times', 10, 'bold'))
self.passkey.grid(row=1, column=1)
# cavana.create_window(100, 100, window=self.passkey)
def loginn():
print('login_test')
admins = {"user_name":"password"}
if self.uname.get() in admins.keys() and self.passkey.get()in admins.values():
master = MasterClass()
window1.destroy()
master.showMain()
else:
label3 = tk.Label(frame, text="user name or password incorrect", font=('times', 10, 'bold'),
fg="red").grid(row=3, column=0, columnspan=2)
button = tk.Button(frame, text="Login", font=('times', 10, 'bold'), padx=20,
command=loginn).grid(row=2,column =0,columnspan =2)
window1.mainloop()
system = Main()
system.login()
### end ###

IJISRT22JUL1344 www.ijisrt.com 874

 Volume 7, Issue 7, July – 2022 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165

Admin login

(login.py) LCD function

(display.py)
Username & LCD Interfa
Keypad function
Admin login password
Keypad (Resistors
User
(resistors) (login.py)
(Keypad.py) LCD function
name Command
Alphanumerical Motor
&pass Username & (display.py)
character interfa
word password LCD Interfa
Data(c
Keypad Admin login
Keypad function ce
Command omma
(resistors) Face (Resistors
(GPIO nds, or
(login.py)Alphanumerical
(Keypad.py) Main Controller
LCD function
data(gross
Username & pins + bits)
User Camera image character
Human Camera pixel)
password (all the detail goes here)
(display.py) Motor
name
face Command drive) LCD Interfa
&pass Keypadmodule (headshot.py)
Activate
Admin login
Keypad function
word (resistors) the camera ) Data(c
Alphanumerical (Resistors
Response(data)
Face
(login.py) character
(Keypad.py) Username & Main Controller
LCD function omma
Human
OTP data(gross Command nds, or
Camera Camera image password
face signal Person’spixel)
ID (all the detail goes here)
(display.py) bits)
User Keypadmodule (headshot.py) Activate command command Motor
Request +LCD
faceInterfa
PIR sensor Response(data)
interfa
name (resistors) Keypad function the camera
Alphanumerical Query data
OTP ce
signal character
Person’sFace
ID result-data (Resistors
&pass
Human (Keypad.py) Main Controller
Camera command command (GPIO Data(c
word
face PIR sensor Face
Camera imagerecognizer data(gross
module Activate (all the detail goes here) pins + +omma
Response(data)
Request face
OTP pixel) Motor nds, or
(recognize.py)
(headshot.py)
signal Database
the function
camera Query databits)
s mobile Person’s ID drive)
command commandresult-data
Human
eviceUser PIR sensor (database.py)
face Camera Main Controller )Response(data)
Query
name
OTP Face
module Camera Face recognizer
image
signal (Keypad.py)
Activate
Person’s
data(gross
ID +command
Request Data(c
face
&pass (all the detail goes here)
the camera
pixel) command command Query dataomma
word PIR sensor (recognize.py)
(headshot.py) result-data DB nds, or
Motor
interfa Data/tokens
Query
s mobile Database function bits)
ce
evice Request +command
face
Face recognizer (database.py) Query (GPIOdata
DB
result-data pins + Data/tokens
(recognize.py) (Keypad.py)
Motor Query
drive) command
Figure 2.2 signal flow grapg DBData/tokens
s mobile )
Face recognizerDatabase function
evice
(recognize.py) (database.py) Query
DB command
Data/tokens
Motor
(Keypad.py) interfa
ce
s mobile
(GPIO
evice Database function pins +
Motor
(database.py)
drive)
(Keypad.py)
)

IJISRT22JUL1344 www.ijisrt.com 875