Professional Documents
Culture Documents
Group 3 - Final Thesis
Group 3 - Final Thesis
Of
Utkarsh Dafe
Nandini Sakhare
Manu Varghese
Pranav Sawarkar
Under the guidance of
Prof. Yogesh Kale (Guide)
Prof. Devendra Shahare (Co-guide)
i
CERTIFICATE OF APPROVAL
Certified that the project report entitled “DEPLOYING A SLIDING GATE BY UTILIZING
COMPUTER VISION AND MACHINE LEARNING TECHNIQUES” has been
successfully completed by UTKARSH DAFE, NANDINI SAKHARE, MANU
VARGHESE, PRANAV SAWARKAR under the guidance of PROF. YOGESH KALE in
recognition to the partial fulfillment for the award of the degree of Bachelor of Engineering in
Electronics and Telecommunication Engineering, Yeshwantrao Chavan College of
Engineering (An Autonomous Institution Affiliated to Rashtrasant Tukadoji Maharaj
Nagpur University).
ii
DECLARATION
We declare that
a. The work contained in this project has been done by me under the guidance of my
supervisors.
b. The work has not been submitted to any other Institute for any degree or diploma.
c. I have followed the guidelines provided by the Institute in preparing the project
report.
d. I have confirmed the norms and guidelines given in the Ethical Code of Conduct of
the Institute.
e. Whenever I have used materials (data, theoretical analysis, figures, and text) from
other sources, I have given due credit to them by citing them in the text of the report
and giving their details in the references. Further, I have taken permission from the
copyright owners of the sources, whenever necessary.
Signature of Students
1) Utkarsh Dafe
2) Nandini Sakhare
3) Manu Varghese
4) Pranav Sawarkar
iii
ACKNOWLEDGEMENT
We also express our sincere gratitude to Principal DR. U. P. WAGHE and DR. M. S.
NARLAWAR, Head of the Department, Electronics and Telecommunication
Engineering, for providing valuable departmental facilities.
We would like to thank PROF. S. S. CHIWANDE (Project Coordinator) and all staff
members of the Electronics Department for extending the facilities without which the
project would not have been a success.
iv
TABLE OF CONTENTS:
CONTENTS PAGE NO
Title Page i
Certificate of Approval ii
Declaration iii
Acknowledgement iv
Table of Content v - vi
List of Figures vi
Abstract vii
Chapter 1: Introduction 1
1.1 Overview
2
1.2 Literature Survey
3
1.3 Problem Statement
4
1.4 Thesis Objectives
4
1.5Thesis Organization
5
v
References
25-26
Societal relevance
27
Co-Po Mapping Table
28
Appendix Publications
28-29
Related Patent search
30
LIST OF FIGURES:
FIGURE PAGE NO
Block Diagram 12
Schematic diagram 12
Aluminum pipe and Angle 13
Pushbutton switch 14
Magnetic Reed switch and Neodymium magnet 14
Voltage Regulator 15
Capacitor,Resistor,Diode and Transistor 15
Node MCU 15
Raspberry Pi and Pi camera 16
vi
Abstract
To maintain security, the majority of individuals use keys, security cards, a
password, or a pattern to unlock the door. The purpose of this research is to assist
persons in increasing the door security of key places through the use of face
detection. Face is a complicated multidimensional structure that necessitates the
use of powerful computational techniques. Face detection, popularly known as
facial recognition, is a computerized technique that uses Computer Vision (CV)
and Machine Learning (ML) (CV and ML are Artificial Intelligence (AI)
specializations) to recognize and validate human faces in collected photos. The
microcontroller is instructed in such a way that the door will automatically open
for the permitted individual. The alert, on the other hand, will sound as a warning
to the unidentified person. This is strictly for the sake of security. We created a
prototype to test our hypothesis and illustrated the team's basic method in this
study piece.
vii
CHAPTER 1
INTRODUCTION
1
1.1 OVERVIEW
Presently, leveraging biometric features instead of chips, codes, or patterns for
computerized personal identification in authentication mechanisms has become
widespread. The entirety of fingerprints data must be retrieved using specialized
equipment such as a fingerprint scanner, palm print scanner, or Genomic analysis
software. Furthermore, during the purposeful sampling step, the objective entities should
come into proximity with the requisite electronics. The initiation step in the system for
facial recognition is face detection. The dependability of face detection impacts the
overall functionality of it. Face detection can distinguish only the visual component of a
photograph impartially of the picture's surroundings. This proposed system gives
complete electronics security with five options; it will reduce the dependency of the user
for redundantly opening or closing the sliding gate. Most of the people install Closed
Circuit Television (CCTV) for security purposes, but their purpose is not solved, the
reason is CCTV is only a vigilance system not security system. This proposed system
gives a complete security system at the main entrance gate of your premises and it gives
all opening and closing possibilities to the user so that the user can control the main
sliding gate on their fingertips. The major goal of this suggested system is to give the user
the ability to manage the main sliding gate internationally via their mobile phone while
also reducing the user's reliance.
The main contribution of this proposed system is to involve the next generation
technologies in the existing manual sliding main gate, it includes Internet of Things (IoT),
image processing and real time processing. These next generation technologies increase
the usage of the main sliding gate, the reason is the usage of all existing manual sliding
gate is not up to the mark, most of the time the complete gate is not open and usually the
user park their two-wheeler and four-wheeler outside of the main sliding gate. This
proposed vehicular entry structure helps to solve the parking problem of many lives.
2
1.2 LITERATURE SURVEY
1)Face Detection
Face detection is one of the most widely used computer vision (CV) for security
applications. In the last decade, multiple face feature detection methods have been
introduced. In recent years, the accomplishment of machine learning and convolutional
neural networks (CNN) have freshly publicized prodigious outcomes in powering
highly-accurate face detection resolutions. Face detection is a computer technique that
controls the position and size of a humanoid face in digital images, although the
suggested approach focuses on the image itself rather than the location. P. Viola and M.
J. Jones used it as the first stage in face recognition is face detection which includes AI,
ML and CV.
The input picture is first converted to grayscale since data processing is much reduced
and it is much easier to operate in a single image plane than a three-dimensional colorful
image without losing vital information. The Viola-Jones (VJ) Algorithm is used to
identify faces. This procedure is quicker than previous methods and produces more
accurate findings.
2)Feature Extraction
For face recognition, N. Dileep Kumar and S. Shanthi utilized the HAAR cascade
methodology. The main reason for selecting this methodology in our system was its real
time face detection capability without any need of specific scale in image size, moreover
it can distinguish both front and side faces. It divides the picture into cells, with each cell
containing pixels that are converted to a binary lighting setup. It will divide the picture
into numerous cells and save them as binary standards.
3
1.3 PROBLEM STATEMENT
❏ If a person wants to cross the gate he/she needs to first park the vehicle and then
slide it, which is rather time-consuming and requires human effort as well.
❏ Security is the most significant element we should have in our home, it gets
compromised whenever any unknown individual attempts to barge in.
A facial recognition and license plate detection sliding gate for security reasons is
designed to improve the security of a limited area by validating the identities of persons
and vehicles attempting to enter.
The face recognition system analyzes facial traits to identify and authenticate individuals'
identities using modern algorithms. This helps to prevent unauthorized access by allowing
only authorized persons to enter.
The license plate detection system, on the other hand, recognises and scans the license
plates of cars approaching the gate before comparing them to a database of authorized
vehicles.
The sliding gate, by combining these two technologies, may effectively limit entry to a
confined area, ensuring that only authorized individuals and vehicles enter.
4
1.5 THESIS ORGANIZATION
Chapter 1 is the official introduction to the thesis, which provides our rationale
for commencing this thesis as well as our goals and objectives for it.
Chapter 3 focuses on our suggested model for this thesis, as well as the workflow
that we followed throughout.
The final findings of our tests and data analysis are presented in Chapter 4.
5
CHAPTER 2
LITERATURE REVIEW
6
2.1 REVIEW OF LITERATURE
[1] R. Chen, "Automatic License Plate Recognition via sliding-window darknet-YOLO deep
learning," Image Vision Computation, vol. 87, pp. 47-56, 2019.
In this paper, they use a You Only Look Once (YOLO)-darknet deep learning system to solve the
problem of automobile license plate detection. We employ YOLO's 7 convolutional layers to
recognise a single class in this research. A sliding-window approach is used for detection. The
goal is to identify Taiwanese license plates. They utilized an AOLP dataset with six-digit license
plate numbers.
[2] K. Bhosale, J. Jadav, S. Kalyankar, R. Bhambare, "Number Plate Recognition System for
Toll Collection," International Journal of Emerging Technology and Advanced Engineering, vol.
4, (4), pp. 729-732, 2014.
In this article, they did a review of the most genuine ways of vehicle license plate detection and an
automatic gate control system that would boost convenience and security at the entry to all key
areas that require protection and security. The gate will operate automatically without the need for
human intervention, and the system will be able to recognise license plates from cars at the entry
gate and determine whether or not to allow vehicles within. The system comprises a Raspberry Pi
and a video camera that takes and analyzes video frames that include a picture of the vehicle
license plate. Python and optical character recognition (OCR) were used to create the suggested
system.
[3] S. S. Farfade, M. J. Saberian and L. Li, "Multi-view face detection using deep convolutional
neural networks," in Proceedings of the 5th ACM on International Conference on Multimedia
Retrieval, 2015, pp. 643-650.
They present Deep Dense Face Detector in this research, a method that does not require landmark
annotation and can recognise faces in a variety of orientations using a single model based on deep
convolutional neural networks. The suggested technique is simple to implement; unlike other
current deep learning object recognition methods, it does not need segmentation, bounding-box
regression, or SVM classifiers. Furthermore, we examined the proposed face detector's scores for
faces in various orientations and discovered that 1) the proposed method can detect faces from
different angles and can handle occlusion to some extent, and 2) there appears to be a correlation
between the distribution of positive examples in the training set and the proposed face detector's
scores.
[4] P. Viola and M. J. Jones, "Robust real-time face detection," International Journal of
Computer Vision, vol. 57, (2), pp. 137-154, 2004.
This research offers a face detection framework that can analyze photos quickly while
obtaining excellent detection rates. The first is the introduction of a novel picture representation
known as the "Integral Image," which allows our detector's features to be calculated very rapidly.
The second is a straightforward and efficient classifier that uses the AdaBoost learning algorithm
to choose a small number of key visual characteristics from a huge pool of possible features. The
third contribution is a method for merging classifiers in a "cascade" that allows background areas
of an image to be swiftly eliminated while more computation is spent on potential face-like
regions.
7
[5] Hteik Htar Lwin, Aung Soe Khaing, Hla Myo Tun, “Automatic Door Access System Using
Face Recognition”, INTERNATIONAL JOURNAL OF SCIENTIFIC & TECHNOLOGY
RESEARCH VOLUME 4, ISSUE 06, JUNE 2015 ISSN 2277-8616.
Hteik Htar Lwin, Aung Soe Khaing and Hla Myo Tun, suggested a method to implement
Automatic Door Access System by directing the flow of information generated by sensors and
cameras through a pc and running the facial recognition modal on MATLAB.
[6] K. Vikram and S. Padmavathi, "Facial parts detection using Viola Jones algorithm," 2017
4th International Conference on Advanced Computing and Communication Systems (ICACCS),
2017, pp. 1-4, doi:10.1109/ICACCS.2017.8014636.
The main motive is to build a system which detects and.recognizes the textures of human parts of
the body in an image or a video. The estimation parameters of the parts in the human body are
tracked with the various parameters of facial features. This can be further applied to real world
applications of face recognition for online exams, identifying persons gender/age, and much more.
[7] Neetu Saini, zukhwinder Kaur, Hari Singh, A Review: Face Detection Methods and
Algorithms, International Journal of Engineering Research & Technology (IJERT), ISSN: 2278-
0181, 2 (6), June – 2013.
Locating and tracking human faces is a prerequisite for face recognition and/or facial expressions.
System needs to capture an image using a camera and frame grabber, process the image, search
the image for important features and then use these features to determine the location of the face.
[8] N. Dileep kumar and S. Shanthi, “Automatic Gate using Face Recognition Technique using
HAAR Cascade Algorithm”, International Journal of Engineering and Advanced Technology
(IJEAT), vol. 9, no. 3, pp. 1302–1305, Feb. 2020, Doi: 10.35940/ijeat.C5195.029320.
Significant work has been carried out in the area of security systems. In this Automatic Sliding
Gate, facial information is recognized and stored in a database. The data processing has been done
by Raspberry Pi. In this, the haar cascade algorithm has been used. This would give more accurate
results compared to Fisher Face, Eigenface method. This will take multiple images in different
angles and check all those images at the time of face recognition.
[9] R. Baran, T. Rusc and P. Fornalski, "A smart camera for the surveillance of vehicles in
intelligent transportation systems," Multimedia Tools Appl, vol. 75, (17), pp. 10471-10493,
2016.
The study describes a smart camera that is intended for security and law enforcement applications
in intelligent transportation systems. The background is supplied in the form of a scholar literature
review initially. In the next sections, the smart camera components and their capabilities for
automatic detection and recognition of specified attributes of automobiles, as well as various
elements of system efficiency, are explained and explored in detail. The system's key advantages
include smart features such as make and model recognition (MMR), licence plate recognition
(LPR), and colour recognition (CR). Their implementations, flowcharts, and recognition rates are
extensively documented, debated, and reported.
8
[10] Oualla, Mohamed & Sadiq, Abdelalim & Mbarki, Samir. (2015). COMPARATIVE
STUDY OF THE METHODS USING HAAR-LIKE FEATURES. INTERNATIONAL
JOURNAL OF ENGINEERING SCIENCES & RESEARCH TECHNOLOGY. 435. 35.
The detection of an item is divided into three stages: the extraction of information that
characterises the object, the processing of this information, and the construction of the final
detector. They will concentrate on the initial stage. Various techniques, such as Haar-Like
characteristics, colour information, texture, edge orientation, and so on, have been used. While
several strategies for extracting information utilising the Haar-like characteristic have been
presented, we are not aware of any surveys on the subject. As a result, this paper presents a
comparative assessment of several studies on Haar-Like characteristics.
[11] H. J. Lee, I. Ullah, W. Wan, Y. Gao, Z. Fang, "Real-time vehicle make and model
recognition with the residual Squeeze Net architecture," Sensors, vol. 19, (5), pp. 982, 2019.
They investigate a few-shot learning issue with an application to vehicle make and model
recognition in traffic surveillance photos in this research. While image-based vehicle
identification systems may be regarded as a more robust alternative to ALPR-based vehicle model
recognition, they frequently fail when an unseen vehicle comes into the picture. Few-shot learning
approaches may provide a potential answer to this unknown class challenge for car model
identification. We compare two common few-shot learning techniques, weight imprinting and
closest class mean classifier, on a vehicle model identification challenge in this paper. The
suggested technique aims to produce strong classification performance on novel classes while
maintaining the high accuracy rate of the base (existing) classes.
[12] Z. Dahirou, M. Zheng and M. Yuxin, "Face Detection with Viola Jones Algorithm," 2020
7th International Conference on Information Science and Control Engineering (ICISCE),
Changsha, China, 2020, pp. 602-606, doi: 10.1109/ICISCE50968.2020.00130.
Face detection is a computer technique that controls the position and size of a humanoid face in
digital images, although the suggested approach focuses on the image itself rather than the
location. P. Viola and M. J. Jones used it as the first stage in face recognition is face detection
which includes AI, ML and CV. The input picture is first converted to grayscale since data
processing is much reduced and it is much easier to operate in a single image plane than a three-
dimensional colorful image without losing vital information. The Viola-Jones (VJ) Algorithm is
used to identify faces. This procedure is quicker than previous methods and produces more
accurate findings.
[13] W. -y. LU and M. YANG, "Face Detection Based on Viola-Jones Algorithm Applying
Composite Features," 2019 International Conference on Robots & Intelligent System (ICRIS),
Haikou, China, 2019, pp. 82-85, doi: 10.1109/ICRIS.2019.00029.
In this study, they suggest using composite features based on the Viola-Jones algorithm to address
the aforementioned difficulties, and they demonstrate the practicality of this strategy through tests.
9
CHAPTER 3
WORK DONE AND DETAILS
10
3.1 TOOLS USED
Hardware:
Chop Saw, Drilling Machine, Drill Bits, Center Punch, Plier, Hammer, C Clamp, etc..
Iron, PCB etching machine, FeCl3 Solution, PCB Drilling Machine, Soldering Iron,
etc..
Software:
Arduino IDE, easyEDA, Raspbian OS, Raspberry Pi Imager, VNC Viewer, etc..
Aluminum Square Pipes, Angles are used for constructing the frame. The Rolling wheels are
used for the frictionless sliding of the gate.
The plastic Rack and Pinion are installed with a geared DC motor on the gate. This is the
main mechanism which drives the gate without human efforts.
To detect the status of the gate, Reed switches are installed with a magnet on the gate. Push
button switches are fixed for the user input
4)Designing of PCB
11
b
Fig. 3.2.1 Block Diagram of The Controller
12
5) Programming the Microcontroller
The Microcontroller is programmed using the open source libraries like BLYNK, esp8266,
etc.
The raspberry pi is interfaced with the pi camera module. Face detection system detect the
face and process it with the data set and commands the microcontroller
The nylon Rack and pinion gear are used for building
the mechanical mechanism which will deploy the
sliding gate. It converts the circular motion from the
Motor to the Linear motion of the gate. The
dimensions of the rack are ½ X 12 inches and it is
super glued over the gate. The Pinion gear has 1 inch diameter and thickness of ¾
inches.
13
Fig. 3.3.4 Relay
The Pushbutton switches are used for the user’s input. The
push buttons are of NO (normally open) type and have a pullup
resistor to normally provide a HIGH signal. When the button is
pressed it provides a LOW output, which is read and processed
by the microcontroller.
14
Fig. 3.3.8 Voltage Regulator
The General Purpose diode 1N4007 is used for the reverse voltage protection from the
coil of relays.
15
Fig. 3.3.11 Raspberry Pi and Pi Camera
The ML and CV processing is done in the Raspberry Pi and the GPIO pins of the
Raspberry pi provide the output to the relay driver to operate the gate.
16
3.4 COSTING
Table 3.1: List of Components, their specifications and Estimated Costing
No. (Approx.)
1) Raspberry Pi – 1 10000
2) Pi Camera – 1 350
3) Pi Case – 1 300
6) Relay 12v 4 60
7) Transistor BC547 5 10
9) Diode 1N4007 10 10
17
CHAPTER 4
18
4.1 RESULT
Object Detection:
Our first objective was to accurately detect humans and number plates within the field
of view of the gate's camera. We utilized a pre-trained ML model which uses Viola
Jones algorithm and OCR (Optical Character Recognition) to achieve this goal.
Object Recognition:
Once objects were detected, our next objective was to recognize them in real-time.
We utilized the AdaBoost algorithm for speeding up the face recognition process.
Finally, our goal was to utilize the object tracking information to control the sliding
gate.
In the proposed system, real time face detection and recognition is done where the
image is matched with the predefined database. If a match is foundthen it outputs
“True” which is interpreted by the microcontroller as to open the gate.
The license plate is detectedand the number is obtained in the form of a text input
which is further checked with the database and if a match is found, then the gate is
opened.
Overall, our project demonstrates the potential of utilizing these technologies for other
applications, such as security systems.
19
Initial face recognition Simulation result
20
License Plate detection:
21
CHAPTER 5
SUMMARY AND CONCLUSION
22
5.1 SUMMARY
A facial recognition and licence plate detection system in a sliding door mechanism is a high-
tech security system that use computer vision algorithms to identify and verify people when
they enter a building or restricted area. A camera module in the system records real-time
video footage of individuals and vehicles approaching the sliding door. The camera data is
then analysed using deep learning algorithms, which allow the system to reliably detect and
recognise people's faces and car licence plates.
The device is designed to function in real time and deliver accurate findings even under
difficult lighting and weather circumstances. The sliding door mechanism is outfitted with
sensors that detect the presence of persons and vehicles, and the facial recognition and licence
plate detection system commences the authentication procedure automatically. To improve
the overall security of the building or restricted area, the system can be combined with other
security systems such as alarms and access control systems.
As a whole, the facial recognition and licence plate detection system in a sliding door
mechanism provides an effective and dependable security management solution, allowing for
rapid and precise identification of persons and vehicles while prohibiting unauthorised
entrance.
5.2 CONCLUSION
We got to work on the project as a group. With more research, ideas, resources and financial
planning , there is much room for improvements and upgrading the system. For the future
scope , addition of number plate detection is definitely one. This way, the vehicle owner
never has to move out of their vehicles to gain access. The proposed system is versatile and
can be used in various types of places like housing society, apartments etc. reducing human
intervention. Apart from the security the system provides flexibility of time too as no one has
to be woken up to open up gates during night times.
23
5.3 FUTURE SCOPE
In the future, the sliding door mechanism's facial recognition and license plate
detection technology might be improved and connected with other technologies to
give a more smooth and efficient user experience.
The application of artificial intelligence (AI) to increase the accuracy and speed of
the facial recognition and license plate identification system is one possible area of
research. This might entail teaching the system to recognise a broader range of face
traits and license plate patterns, as well as adjust to changing lighting and climatic
circumstances.
Furthermore, the system might be linked to a home automation platform, allowing
users to remotely manage the door and monitor entry using a smartphone app or
other device.
Finally, the facial recognition and license plate detection system might be modified
to operate with doors and access points other than sliding doors. Automatic gates,
garage doors, and other sorts of access points that might benefit from better security
and access control capabilities could be included.
24
REFERENCES:
[3] S. S. Farfade, M. J. Saberian and L. Li, "Multi-view face detection using deep
convolutional neural networks," in Proceedings of the 5th ACM on International
Conference on Multimedia Retrieval, 2015, pp. 643-650.
[4] P. Viola and M. J. Jones, "Robust real-time face detection," International Journal
of Computer Vision, vol. 57, (2), pp. 137-154, 2004.
[5] Hteik Htar Lwin, Aung Soe Khaing, Hla Myo Tun, “Automatic Door Access
System Using Face Recognition”, INTERNATIONAL JOURNAL OF SCIENTIFIC
& TECHNOLOGY RESEARCH VOLUME 4, ISSUE 06, JUNE 2015 ISSN 2277-
8616.
[6] K. Vikram and S. Padmavathi, "Facial parts detection using Viola Jones
algorithm," 2017 4th International Conference on Advanced Computing and
Communication Systems (ICACCS), 2017, pp. 1-4,
doi:10.1109/ICACCS.2017.8014636.
[7] Neetu Saini, zukhwinder Kaur, Hari Singh, A Review: Face Detection Methods
and Algorithms, International Journal of Engineering Research & Technology
(IJERT), ISSN: 2278-0181, 2 (6), June – 2013.
[8] N. Dileep kumar and S. Shanthi, “Automatic Gate using Face Recognition
Technique using HAAR Cascade Algorithm”, International Journal of Engineering
and Advanced Technology (IJEAT), vol. 9, no. 3, pp. 1302–1305, Feb. 2020, Doi:
10.35940/ijeat.C5195.029320.
[9] R. Baran, T. Rusc and P. Fornalski, "A smart camera for the surveillance of
vehicles in intelligent transportation systems," Multimedia Tools Appl, vol. 75, (17),
pp. 10471-10493, 2016.
[10] Oualla, Mohamed & Sadiq, Abdelalim & Mbarki, Samir. (2015).
COMPARATIVE STUDY OF THE METHODS USING HAAR-LIKE FEATURES.
INTERNATIONAL JOURNAL OF ENGINEERING SCIENCES & RESEARCH
TECHNOLOGY. 435. 35.
[11] H. J. Lee, I. Ullah, W. Wan, Y. Gao, Z. Fang, "Real-time vehicle make and
model recognition with the residual Squeeze Net architecture," Sensors, vol. 19, (5),
pp. 982, 2019.
25
[12] Z. Dahirou, M. Zheng and M. Yuxin, "Face Detection with Viola Jones
Algorithm," 2020 7th International Conference on Information Science and Control
Engineering (ICISCE), Changsha, China, 2020, pp. 602-606, doi:
10.1109/ICISCE50968.2020.00130.
26
SOCIETAL RELEVANCE:
Face recognition and licence plate detection technologies integrated with sliding
door mechanisms can have various societal implications, including:
1) Enhanced Security:
The facial recognition and licence plate detection technology may be used to
prevent unauthorised access to restricted places including businesses, hospitals,
and schools. This will assist to strengthen security and lessen the danger of theft
or vandalism at these locations.
2) Contactless entry:
3) Traffic Management:
Number plate recognition systems may be used to track the number of cars
entering and exiting a facility, which can help with traffic management and
planning.
5) Crime Prevention:
The licence plate detection system may be used to identify and monitor cars
involved in criminal activity. This helps to prevent and identify crimes like
robbery and auto theft.
27
CO-PO MAPPING TABLE:
APPENDIX:
Publication
28
29
RELATED PATENT SEARCH:
https://patents.google.com/patent/US9527115B2/en
https://patents.google.com/patent/KR20160040037A/en?q=(face+recognition+securit
y)&oq=face+recognition+for+security
https://patents.google.com/patent/US10438053B2/en?q=(face+recognition+security)
&oq=face+recognition+for+security
https://patents.google.com/patent/TW202213237A/en?q=(face+recognition+security)
&oq=face+recognition+for+security
https://patents.google.com/patent/US10930151B2/en?q=(license+plate+detection+ent
ry)&oq=license+plate+detection+for+entry
30
Yeshwantrao Chavan College of Engineering
Project Preliminary Investigation Report
Name of Department:
Electronics and Telecommunication Engineering
Students Details:
Roll No. Name of Student Email ID Mobile No.
31
Problem Statement:
Designing a face recognition and license plate detection system for a sliding door
mechanism that ensures secure access control in a commercial or residential setting.
The sliding door mechanism is a common feature in many buildings, and it is essential to
ensure that only authorized individuals are granted access. The primary challenge is to
develop a system that can accurately recognize individuals and detect license plates in real-
time.
The face recognition and license plate detection system must be able to handle different
lighting conditions, angles, and variations in appearance, such as facial hair, glasses, and
hats. It must also be able to detect fake faces and license plates that may be used to gain
unauthorized access.
Security is a major concern, and the system must be designed to prevent hacking or
unauthorized access to the system's data. The system must ensure that the data collected is
secure and protected from cyber attacks, as well as physical attacks.
To ensure the security of the system, it is essential to implement strict access control
policies, such as limiting access to only authorized personnel and ensuring that all access
attempts are logged and monitored. The system should also be designed to be tamper-proof
and resistant to physical attacks.
Overall, the development of a face recognition and license plate detection system for a
sliding door mechanism is crucial in ensuring the security and safety of individuals and
property. The system must be designed to be accurate, reliable, and secure, while also being
user-friendly and easy to operate.
32
Prior Art (Patent Search):
Patent Title of Patent Existing Solutions
Application No. (Abstract of Patent)
Computer vision and The present invention encompasses software
US9527115B2 machine learning that brings together computer vision and
software for grading machine learning algorithms that can
and sorting plants evaluate and sort plants into desired
categories. While one embodiment of the
present invention is directed toward
strawberry plants, the software engine
described is not specifically designed for
strawberry plants but can be used for many
different types of plants that require
sophisticated quality sorting. The present
invention is a sequence of software
operations that can be applied to various
crops (or other objects besides plants) in a
reusable fashion.
33
US10438053B2 Biometric An exemplary embodiment of the present
identification invention provides a method of verifying an
systems and identity of a person-to-be-identified using
methods biometric signature data. The method
includes creating a face sample database
based on biometric signature data from a
plurality of individuals, calculating a feature
database by extracting selected features of
entries in the sample database, calculating
positive samples by calculating a feature
absolute value distance for a same position
of any two different images from one
person, calculating negative samples by
calculating a feature absolute value distance
for a same position of different people,
calculating a key bin feature using a learning
algorithm, calculating a classifier from the
key bin feature for use in identifying and
authenticating an acquired face image of a
person-to-be-identified and identifying and
authenticating the person-to-be-identified
using the classifier and the acquired face
image of the person-to-be-identified.
34
US10930151B2 Roadside parking A roadside parking management method
management includes: receiving a monitoring area image
method, device, captured by a camera; dividing the
and system based monitoring area image to obtain a plurality
on multiple of tracking detection areas in the monitoring
cameras area image; monitoring at least one of a
vehicle and a license plate in the monitoring
area image; and determining parking event
information according to a tracking
detection area where the at least one of the
vehicle and the licence plate is located. The
method can improve the management
efficiency of parking events.
Literature Review:
Title of Paper Details of Literature Identified for
Publication with Project
Date and Year
“Automatic License Plate Vol. 87, pp. 47-56, 2019 Image Vision Computation
Recognition via sliding-
window darknet-YOLO deep
learning”
“Number Plate Recognition Vol. 4, (4), pp. 729-732, International Journal of Emerging
System for Toll Collection” 2014 Technology and Advanced
Engineering
35
Current Limitations:
A facial recognition and licence plate detection system in a sliding door mechanism
currently has various limitations:
(1) Accuracy:
Lighting conditions, distance from the camera, facial expression, and angle of the face can
all impact the accuracy of face recognition and licence plate identification. As a result, the
system may not always correctly identify persons or licence plates, thereby posing a
security risk.
The use of face recognition technology creates privacy concerns since it entails the
collection and storage of biometric data. People may be concerned about their faces being
scanned and saved, particularly if they are unaware that this is occurring.
The system can generate false positives when it mistakenly detects a person or licence
plate, and false negatives when it fails to recognise a person or licence plate. These
mistakes can cause annoyance, delays, and security breaches.
Only facial recognition and licence plate detection are currently supported by current
technology. It cannot distinguish between authorised and unauthorised people or vehicles,
nor can it provide extra security features like biometric authentication or motion detection.
36
Proposed Solution:
The following components might be included in a suggested solution for a facial
recognition and licence plate detection system in a sliding door mechanism:
The initial component of this approach is a camera system capable of capturing crisp
photos of faces and licence plates. The camera should be positioned such that it can see the
sliding door mechanism's entrance and exit.
The second component of the system is facial recognition software, which can identify and
match the faces of people entering and departing the sliding door mechanism. Face
recognition software should be able to recognise faces even in low-light settings and from
various angles.
The third component of the solution is licence plate recognition software, which detects and
reads vehicle licence plates as they enter and depart the sliding door mechanism. Even if
the licence plate is dusty, broken, or partially concealed, our programme should be able to
recognize it.
The fourth component of the solution is a control system that manages the sliding door
mechanism's opening and shutting depending on the findings of the face and licence plate
recognition software. This system should be able to recognise authorised people and
vehicles and enable them to enter and depart the sliding door mechanism.
(5) Database: The solution's fifth component is a database that holds the authorised faces
and licence plate numbers. The facial and licence plate recognition software may utilise this
information to authenticate the identities of persons and vehicles entering and departing the
sliding door mechanism.
37
Objectives and Scope of Work:
Objectives:
A facial recognition and licence plate detection system in a sliding door mechanism may have
the following goals:
Scope of Work:
38
Feasibility Assessment:
(1) A facial recognition and licence plate detection system in a sliding door mechanism is
projected to result in a more secure and easy access control system for a building or facility.
(2) The system would be able to recognise authorised persons and let them to enter the facility
without the use of physical keys or access cards using face recognition technology. This would
make authorised persons' access to the building more efficient and easy, while simultaneously
boosting security by prohibiting unauthorised ones from getting access.
(3) Similarly, using licence plate recognition technology, the system could automatically identify
and authorise cars permitted entrance to the premises. This might be particularly beneficial for
parking garages or gated facilities, allowing for smooth entry and departure without the need for
drivers to roll down their windows or scan access cards.
(4) Overall, including facial recognition and licence plate detection technology into a sliding
door mechanism would result in a more efficient and secure access control system that would
allow authorised personnel and vehicles to enter and depart more quickly.
(5) Customization:
The system may be tailored to each building's individual requirements, allowing for features such
as automatic visitor check-ins or automatic lockdown processes in the case of a security breach.
Overall, incorporating facial recognition and licence plate detection technologies into sliding door
systems has the potential to transform building security and access control. This innovation has
the potential to become a key component of modern building management due to its capacity to
increase security, efficiency, and give useful data for study.
39
III. Tasks Involved:
Some of the jobs involved in a facial recognition and licence plate detection system in a sliding
door mechanism are as follows:
40
IV. Expertise Required:
The following skills are necessary to create a facial recognition and licence plate detection
system in a sliding door mechanism:
(6) Security:
A thorough grasp of security is required to guarantee that the facial recognition and licence
plate detection system is safe and secure from hackers and unauthorised access.
41
V. Facilities Required:
The following facilities may be required to integrate a facial recognition and licence plate
detection system into a sliding door mechanism:
(1) Camera:
High-quality cameras are required to capture good photos of faces and licence plates. To achieve
reliable identification, the camera should have a high resolution and strong low-light capability.
To analyse the images taken by the camera and extract the data required for facial recognition and
licence plate detection, specialised image processing software is required.
(3) Database:
A database is needed to hold the images and data gathered by the camera for comparison and
identification. The database should be capable of handling massive volumes of data and have
effective search engines.
(4) Connectivity:
To enable remote monitoring and control, the system should be linked to a network, such as the
internet or a local area network.
The sliding door mechanism should be powered by a dependable power source capable of handling
the increased demand imposed by the facial recognition and licence plate detection systems.
Because this is a security system, it is critical that it has strong security measures in place to
prevent unauthorised access and data breaches.
The system should be built to function in sync with the motions of the sliding door mechanism.
To engage with the system and access its capabilities, such as monitoring and operating the door
and reviewing recorded data, users require an intuitive and user-friendly interface.
42
Milestones and Time Plan:
Task J A S O N D J F M A
U U E C O E A E A P
L G P T V C N B R R
2022 2022 2022 2022 2022 2022 2023 2023 2023 2023
Desig Conceptual ✔
n Design
Detailed ✔
design
Design ✔
Modification
Final Design ✔
Devel Procurement ✔ ✔
op (If any)
Prototyping ✔ ✔
Modification ✔
Final ✔ ✔
Modification
IPR / patent ✔ ✔
draft
Thesis and ✔ ✔
Poster
43