2021 International Conference on Advance Computing and Innovative Technologies in Engineering (ICACITE)

Department of Electrical & Electronics Engineering, Galgotias College of Engineering and Technology, Gr. Noida, India

License Plate Recognition System

2021 International Conference on Advance Computing and Innovative Technologies in Engineering (ICACITE) | 978-1-7281-7741-0/20/$31.00 ©2021 IEEE | DOI: 10.1109/ICACITE51222.2021.9404706

Farheen Ali Himanshu Rathor Wasim Akram

Galgotias University University, Galgotias University University, Galgotias University University,
GreaterNoida GreaterNoida GreaterNoida

Abstract - The most successful and cost-effective method II. PROBLEM FORMULATION
used for car identifying purposes is detection of the License
Plate. These approaches and methods differ depending on In several nations, such as the United States of America,
factors such as image quality, the vehicle at fixed locations, Australia, in Korea and a few others, license plate
light conditions, single image, etc. The differences in license recognition programs have been introduced. The early
plates from various nations and states should also be able to development of license plate recognition systems has been
deal with it. The method should also be able to work smoothly assisted by the stringent enforcement of license plate
with the number of characters in the captured images that requirements in these countries. These systems use
differ in plate size or plate size. traditional license plate features such as plate height, plate
boundary, character color, and font, etc. These systems help
This product-oriented project involves the design and to accurately find the number plate and recognize the
development of a License Plate recognition software system vehicle's license number.
that would be useful in vehicle identification as well as in-
vehicle parking services and even e challan. We will Number plate norms are scarcely observed in India. In
concentrate primarily on identifying and recognizing multiple terms of font styles, script, scale, location, and color of the
license plate cars from a single frame. The proposed system number of plates, large variations are found. In a few
consists of two steps: Identification of plate numbers and instances, the number plate includes other unwelcome
identification of them. decorations. Unlike other nations, there are also no special
features present on Indian numbers, plates to ease their
The number plate is identified from the picture taken in the process of identification.
plate number detection process and then the segmented plate is
transferred to the plate recognition in the second phase to The overall goal of this research is to establish an Indian
determine the characters and numbers. License Plate Identification deployable working method,
which is a difficult challenge since there is no uniformity of
Key-Words - Recognition, Captured Image, Segmented Plate, the Indian License Plates since each state has followed its
License Plate, Identification, E- Challan. own design and structure. In addition to the challenges found
by the General License Plate Recognition Algorithms, the
I. INTRODUCTION Indian Algorithms for License Plate Recognition need to fix
License Plate Recognition is an image processing system some complex problems.
that recognizes the car using the number plate. With the use As the Indian Transportation Engineering and
of the vehicle number plate, the aim is to design, an effective Infrastructure is insufficient, the road rules are ignored and
automatic approved vehicle identification system. The the number of injuries is enhanced by the lack of surveillance
system is introduced for safety monitoring at the entry to a capabilities. Identifying cars using their license plates is the
highly confined region, such as combat zones or areas primary need. But the Indian license plates and facilities are
surrounding top government departments, e.g. Parliament, not standardized.
the Supreme Court, etc.
Depending on the car class, the license plates can’t be
Each individual may use this system for security placed in the right position. This will lead to the plate image
purposes. An android program will be mounted on your cell in the vehicle and the vehicles being misclassified. The
phone for the general public. After that, if he needs to know appearance of odd characters can be misclassified as
the specifics of any car, he only has to catch the license plate Number Plate Characters in the License Plate.
image and then process that image to get the information he
wants about that vehicle. This system is really critical and The non-uniformity of license plates depends on the car type.
needs to be introduced.
In the License Plate, single / multi-line characters.
The developed device senses the vehicle for defensive
Dust / Unused License Plates.
purposes first and then captures the picture of the vehicle.
Non-Uniformity present in Font Types, Font Size and Color.
By segmenting the picture into an image, the region of
the vehicle number plate is extracted. The Optical Character The concerns to be resolved by the above specification
Recognition Technique (OCR) is used for character needs to be solved by an appropriate framework.
recognition. The resulting data is then used to compare the
records on a database to provide unique information such as
the identity of the owner, registration location, address, etc.
The framework is implemented and simulated in JAVA, and
it is tested on actual picture results.

978-1-7281-7741-0/21/$31.00 ©2021 IEEE 1053

Authorized licensed use limited to: University of Technology Sydney. Downloaded on May 22,2021 at 00:50:04 UTC from IEEE Xplore. Restrictions apply.
 2021 International Conference on Advance Computing and Innovative Technologies in Engineering (ICACITE)
Department of Electrical & Electronics Engineering, Galgotias College of Engineering and Technology, Gr. Noida, India

Frontend – HTML, JavaScript,

Android (for developing an android app), CSS,
Bootstrap.
Applications – PyCharm, Android Studio, VSCode.
VI. LITERATURE REVIEW
There are many research done in past on this topic some
of them are as follows:
A. 1.Using Text-line Construction and Multilevel RBF
Neural Network:
III. MERITS OF PROPOSED SYSTEM In this method, the author used a latest method

x It is based on android app. for vehicle plate recognition using text-line construction
and multilevel RBFNN. during this method localization of
x It would work with less expensive cameras. the vehicle plate are going to be done using the text-line
x Does not require infrared lighting or sensors as construction result and also the characteristics of the
system camera is enough. character arrangement.
x It works in real time.
Then binarization is employed to urge the more accurate
x Easily accessible because it is based on android. location of vehicle plate. then, the strategy of vertical
x It is specially made for Indian conditions. projection information with prior knowledge is proposed to
x It is fast and accurate. segment character and extract the statistic feature. Then
feature vectors are recognized using multilevel classification
x It is cost effective.
RBF neural network.
x Easy to use by anyone.
x User friendly Interface. B. 1.Using Radial Basis Function Neural Network:
In this method, the author used Radial Basis Function.
IV. WORK PLAN Neural Network (RBFNN). Other pre- processing processes
1) Designing a Front End UI which is user friendly and like edge detection, image dilation, filtering, filling and
easy to manageand use. As per thedemand. smoothing also are used for improving image quality. In pre-
processing first image is converted to gray scale image, then
2) Training our Machine Learning Model with highest
into black and white using sobel edge detector and
precision so that detection is as close as possible morphological method.
toaccuracy.
3) Coding and using the already made Machine After this the pre-processed image is employed for
Learning Model in the OpenCV with the help vehicle plate recognition using threshold value, the objects
those satisfy the most and minimum threshold values are
ofpython.
then classified as “Plate” and “No Plate” categories. Then
4) Integrating frontend and backend as per the user or RBFNN is employed for detection of vehicle plate the
device requirements. objects those are classified as “Plate” are finally used for
5) Testing and fixing any bugs detected. trained RBFNN.
6) Upgrading for any possible issues or
performanceproblems.
7) Work on the feedback received fromcustomers.
8) Results

VII. GRAPHICAL USER INTERFACE

The interface of the app is very simple and simple to use.
The user doesn't got to take any additional things to know the
functionality of the app.
The UI is formed in such how that user can easily
understand where the input needs to be given. Steps are
given to assist the user in giving the right input.
V. TOOLS
The following are main things which appear on the
Backend – Python, OpenCV, Machine Learning. screen.
x Home Page

1054

Authorized licensed use limited to: University of Technology Sydney. Downloaded on May 22,2021 at 00:50:04 UTC from IEEE Xplore. Restrictions apply.
 2021 International Conference on Advance Computing and Innovative Technologies in Engineering (ICACITE)
Department of Electrical & Electronics Engineering, Galgotias College of Engineering and Technology, Gr. Noida, India

x Capture Image page solution of this problem, with respect to the cities and
countries where it is in used.
x Text page
Many developers around the world are now working on
C. Result page
LPR system that gives high-quality footage with extreme
Advantages. accuracy in any country around the world where its is in
used. Most challenges related to LPR are solved and the
1) Monitoring and Surveillance.
concern related to the collected data should be taken care.
With the help of license plate recognition, the stressful,
long and time-consuming process of recording numbers XII. CONCLUSIONS
manually will be avoided easily. It's impossible to spot the
right license plate number, especially during bad weather or The main purpose of this research paper is to investigate
when a car is running very fast. Along with recording license the possibilities of getting all the details related to vehicle
plate numbers, LPR systems can give a clear insight to traffic using license plate recognition. To improve the quality of the
whereabouts in real-time. vehicles images we can use fusion technique, then extract the
license plate and isolate the characters which are present on
2) Better Security the number plate and then identify the characters on the
License number plate recognition is capable to deter license plate using artificial neural network. The
constant traffic offences. It is very helpful to security like experimental have shown that the ability to recognize the
inspections, investigations, and legal matters. Number plate characters correctly on license plate by neural network has
recognition very helpful especially in a parking lot or at probability of 95% even in presence of noise with 50%
places where there are heavy traffics. According to the density. This proposed approach of LPR can be implemented
survey, it is very useful to catch stolen vehicles using by police to detect speed violators, no-parking zone
number plate recognition app. voilators, parking areas, highways, bridges or tunnels. Also
the prototype of the system is going to be integrated and
3) Cost-effective
tested as part of the sensor network being developed by
Since it is easier and more efficient, LPR technology is
Ayers Island, LLC as part of their Intelligent Island system.
also one of the most cost-effective solutions for managing
your car park. It will become very easy to cut costs and XIII. REFERENCES
lower the need for security personnel when you use this
[1] APNR Survey :
smart technology. Traffic police will easily issue fines to https://www.researchgate.net/publication/236888959_Automatic_Nu
anyone picked up by their LPR app that shouldn’t be in No- mber_Plat e_Recognition_System_ANPR_A_Survey
Parking area or anyone who has exceeded the maximum [2] APNR for Vehicle Identification using OCR:
speed limit. This will bring in extra money for the traffic https://www.researchgate.net/publication/224565257_Automatic_Nu
police. mber_Plat
e_Recognition_System_for_Vehicle_Identification_Using_Optical_
4) Self-sufficient. Character_ Recognition
LPR cameras can operate in such a way such that all [3] Wikipedia: https://en.wikipedia.org/wiki/Automatic_number-
information related to vehicles is completely processed so plate_recognition
that we don't need any extra computers or software. LPR [4] Image 01: https://www.pyimagesearch.com/2020/09/21/opencv-
automatic- license-number-plate-recognition-anpr-with-python/
cameras have optical character recognition software
embedded which helps in analysing all the images directly. [5] Image 02: https://www.researchgate.net/figure/License-Plate-
Recognition- System-Phases_fig1_272863336
Due to being self-sufficient solutions, LPR cameras are
[6] Image 03: http://www.solutionshead.in/industry-
quick, safe and light to install. solutions/management- solutions/automatic-number-plate-
recognition-system
Limitations:
[7] Book: Vehicle License Plate Recognition: A Soft Computing Based
1) Privacy Concerns Approach by Laxmi Vijaya (Author), Das Madhulika (Author)
People's are worried about their privacy as LPR record
images and stored them . People's data data can be stolen by
hacker or people with nefarious intentions. But this LPR isn’t
stealing someone’s privacy. LPR is way more efficient and
time-saving than a manual process because it checks every
vehicle plate automatically. These privacy related issues are
into consideration so that people can feel safer on the road
instead of worrying about their privacy.
2) Extreme Weather Conditions
During bad weather conditions license plate recognition
systems is not very much effective.So in that case all the
security measures will be shut down and manual surveillance
will be needed.
XI. CHALLENGES LPR SOFTWARE FACES
There are some problems that license plate recognition
face and non-uniformity of number plates for different
towns, cities and countries is the most common challenge
that LPR face.In some cases length of number plates also
vary.The LPR must be customized to come up with the

1055

Authorized licensed use limited to: University of Technology Sydney. Downloaded on May 22,2021 at 00:50:04 UTC from IEEE Xplore. Restrictions apply.