Journal of University of Shanghai for Science and Technology ISSN: 1007-6735

Recognition of Vehicle Number Plate Using Matlab

Dr. B. Mouli Chandra, M.Tech.,Ph.D.,1, D Sonia2, A Roopa Devi3, Ch Yamini Saraswathi4, K
Mighty Rathan5, K Bharghavi6
1 Professor& HOD, Department of EEE, QIS College of Engineering and Technology, Ongole.
2,3,4,5,6 B.Tech Scholars, Department of EEE, QIS College of Engineering and Technology, Ongole.

Abstract:
Video surveillance system is used for of numbers and characters present on
security purpose as well as monitoring number plate, template matching approach is
systems. But Detection of moving object is a used to recognition of numbers and
challenging part of video surveillance. characters. The concentrate is given to
Video surveillance system is used for Home locate the number plate region properly to
security, Military applications, Banking segment all the number and letters to
/ATM security, Traffic monitoring etc. Now identify each number separately.
a day’s due to decreasing costs of high
quality video surveillance systems, human Introduction
activity detection and tracking has become Number plates are used for identification of
increasingly in practical. Accordingly, vehicles all over the nations. Vehicles are
automated systems have been designed for identifying either manually or automatically.
numerous detection tasks, but the task of Automatic vehicle identification is an image
detecting illegally parked vehicles has been processing technique of identify vehicles by
left largely to the human operators of their number plates. Automatic vehicle
surveillance systems. The detection of identification systems are used for the
Indian vehicles by their number plates is the purpose of effective traffic control and
most interesting and challenging research security applications such as access control
topic from past few years. It is observed that to restricted areas and tracking of wanted
the number plates of vehicles are in different vehicles. Number plate recognition (NPR) is
shape and size and also have different colour easier method for Vehicle identification.
in various countries. This work proposes a NPR system for Indian license plate is
method for the detection and identification difficult compared to the foreign license
of vehicle number plate that will help in the plate as there is no standard followed for the
detection of number plates of authorized and aspect ratio of license plate. The
unauthorized vehicles. This paper presents identification task is challenging because of
an approach based on simple but efficient the nature of the light. Experimentation of
morphological operation and Sobel edge number plate detection has been conducted
detection method. This approach is from many years; it is still a challenging
simplified to segmented all the letters and task. Number plate detection system
numbers used in the number plate by using investigates an input image to identify some
bounding box method. After segmentation local patches containing license plates. Since

Volume 23, Issue 2, February - 2021 Page-363

Journal of University of Shanghai for Science and Technology ISSN: 1007-6735

a plate can exist anywhere in an image with License plate location algorithm consist of
various sizes, it is infeasible to check every steps like as Edge Detection, Morphological
pixel of the image to locate it. In parking, operation like dilation and erosion,
number plates are used to calculate duration Smoothing, segmentation of characters and
of the parking. When a vehicle enters an recognition of plate characters are described.
input gate, number plate is automatically Images and Digital Images:
recognized and stored in database. In NPR
system spectral analysis approach is used
were acquiring the image, extract the region
of interest, character segmentation using
SVM feature extraction techniques. The
advantage of this approach is success full
recognition of a moving vehicle. It is
difficult to detect the boundary of the
Number plate from the input car images in
outdoors scene due to colour of characters of
the number plate and Background of the
Number plate the gradients of the original
image is adopted to detect candidate number Figure 2 digital & analog images
plate regions. There are also algorithms • A digital image differs from a photo in that
which are based on a combination of the values are all discrete.
morphological operation, segmentation and • Usually, they take on only integer values.
Canny edge detector. • A digital image can be considered as a
large array of discrete dots, each of which
has a brightness associated with it. These
dots are called picture elements, or more
simply pixels.
• The pixels surrounding a given pixel
constitute its neighbourhood A
neighbourhood can be characterized by its
shape in the same way as a matrix: we can
speak of a 3x3 neighbourhood, or of a 5x7
neighbourhood.

Aspects of Image Processing

Image Enhancement: Processing an image
so that the result is more suitable for a
Figure 1.IMAGE PROCESS particular application. (sharpening or
deblurring an out of focus image,
highlighting edges, improving image

Volume 23, Issue 2, February - 2021 Page-364

Journal of University of Shanghai for Science and Technology ISSN: 1007-6735

contrast, or brightening an image, removing Recognition Of Vehicle Number Plate

noise) Using MATLAB
Image Restoration: This may be considered Ami Kumar Parida, SH Mayuri, Pallabi
as reversing the damage done to an image by Nayk, Nidhi Bharti
a known cause. (removing of blur caused by Automatic video analysis from traffic
linear motion, removal of optical distortions) surveillance cameras is a fast-emerging field
Image Segmentation: This involves based on computer vision techniques. It is a
subdividing an image into constituent parts, key technology to public safety, intelligent
or isolating certain aspects of an image. transport system (ITS) and for efficient
(finding lines, circles, or particular shapes in management of traffic. We define video
an image, in an aerial photograph, analytics as computer-vision-based
identifying cars, trees, buildings, or roads. surveillance algorithms and systems to
extract contextual information from video.
Literature Survey Currently most reliable approach is through
the recognition of number plates, i.e.,
Vehicle Number Plate Recognition automatic number plate recognition
System: A Literature Review and (ANPR), which is also known as automatic
Implementation using Template license plate recognition (ALPR), or radio
Matching Aniruddh Puranic, Deepak frequency transponders. We are proposing
K.T. Umadevi V two methods for extraction of license plates
The growing affluence of urban India has and comparing it with other existing
made the ownership of vehicles a necessity. methods. The Extracted license plates are
This has resulted in an unexpected civic segmented into individual characters by
problem - that of traffic control and vehicle using a region-based method. The
identification. Parking areas have become recognition scheme combines adaptive
overstressed due to the growing numbers of iterative thresholding with a template
vehicles on the roads today. The Automatic matching algorithm
Number Plate Recognition System (ANPR)
plays an important role in addressing these VEHICLE NUMBER PLATE
issues as its application ranges from parking DETECTION USING MATLAB
admission to monitoring urban traffic and to Narendra Singh Tomar, Prakhar Sachan,
tracking automobile thefts. There are Pranav Mittal, Shivani Agarwal
numerous ANPR systems available today The VPR (Vehicle Number plate
which are based on different methodologies. Recognition) system is based on image
In this paper, we attempt to review the processing technology. It is one ofthe
various techniques and their usage. The necessary systems designed to detect the
ANPR system has been implemented using vehicle number plate. In today’s world with
template Matching and its accuracy was the increasing number of vehicle day by day
found to be 80.8% for Indian number plates. it’s not possible to manually keep a record
of the entire vehicle. With the development

Volume 23, Issue 2, February - 2021 Page-365

Journal of University of Shanghai for Science and Technology ISSN: 1007-6735

of this system it becomes easy to keep a recognized. ANPR can be used to store the
record and use it whenever required. The images captured by the cameras and the text
main objective here is to design an efficient from the number plate. Systems use infrared
automatic vehicle identification system by lighting to allow the camera to take the
using vehicle number plate. The system first picture at any time of day. A powerful flash
would capture the vehicles image as soon as can also included in cameras, to both
the vehicle reaches the security checking illuminate the picture and make the offender
area. The captured images are then extracted aware of his mistake. Due to plate variation
by using the segmentation process. Optical from place to place ANPR technology tends
character recognition is used to identify the to be region specific.
characters. The obtained data is then
compared with the data stored in their Implementation
database. The system is implemented and
Npr implementation using matlab:
simulated on MATLAB and performance is
tested on real images. This type of system is Vehicle Image Captured By Camera:
widely used in Traffic control areas, tolling, The image of the vehicle
parking area .etc. This system is mainly whose number plate is to be identified
designed for the purpose of security system. is captured using digital camera of 3.2
megapixel.
Automatic Vehicle Number Plate
Recognition System using Matlab
Bhawna Tiwari, Archana Sharma, Malti
Gautam Singh, Bhawana Rathi
Automatic number plate recognition is a
mass surveillance method that uses optical
character recognition on images to read the
number plates on vehicles. Existing closed-
circuit television or road-rule enforcement
cameras, or specifically designed systems
can be used for the task. This system is very
helpful for traffic police to find the details of
a car violating the traffic rules. Its
applications also includes Automatic toll
collection system and car parking
systems.[1] In high security areas where
parking space is reserved for VIP vehicle
owners only, the parking gate will be opened Extraction Of Number Plate Location:
after number recognition. In areas where
parking space is allotted to a particular RGB to gray-scale conversion is
vehicle, wrong vehicle parked can be adopted, in order to facilitate the plate

Volume 23, Issue 2, February - 2021 Page-366

Journal of University of Shanghai for Science and Technology ISSN: 1007-6735

extraction, and increase the processing wslinesofhighcontrastintheimage.Th

speed. This conversion is used eselinesdonotquitedelineatetheoutlin
I_gray=0.114*R+0.587*G+0.299*B. e of the object of interest. Compared
to the original image, gaps in the
lines are observed that surrounds the
object in the gradient mask. This
linear gap disappears if the Sobel
image is dilated using square
structuringelements.

Then we take the binary image,

binary image is closed using square MATLAB toolbox provide a
structuring elements to facilitate the function imfill (BW, “holes”) that
plate extraction. Mathematical fills holes in the binary image. The
morphology is used to detect the dilated gradient mask shows the
region of interese and Sobel operator outline of the cell quite nicely, but
are used to calculate the threshold there are still holes in the interior of
value, that detect high light regions the cell.
with high edge magnitude and high
edge variance. The binary gradient
masksho

Volume 23, Issue 2, February - 2021 Page-367

Journal of University of Shanghai for Science and Technology ISSN: 1007-6735

segmented object look natural, the image is

eroded twice with diamond and line
structuring element. This helps in extraction
of number plate area of the vehicle.

Remove Connected Objects On Border:

The region of interest has Togettheonlynumberplateare
been successfully segmented, but it is ainavehicleimagewithcharactersandnu
not the only object that has been found. mberspresentonitthesegmented image
Any objects that are connected to the is multiplied with binary image.
border of the image can be removed
using the imclear border MATLAB
function.

Character Segmentation:
Finally, in order to make the Segmentation is one of the most important

Volume 23, Issue 2, February - 2021 Page-368

Journal of University of Shanghai for Science and Technology ISSN: 1007-6735

processes in the number plate recognition, of manual entry. This project was a success
because all further steps rely on it. If these in recording the number plate ofa vehicle
gmentation fails, a character can be although it has got its own limitation of
improperly divided into two pieces, or two image processing and other hardware
characters. The ultimate solution on this requirements.
problem is to use bounding box technique.
Once a bounding box created over each References
character and numbers presented on number
[1]T. Pratheeba, “Morphology Based Text
plate, each character & number is separate
Detection and Extraction from Complex
out for recognition of number plate.
Video Scene," International Journal of
In the bounding box technique we used Engineering and Technology Vol.2(3),
some instructions, bw label to detect the 200-206,2010.
number of connected elements in the image
and the matrix of the image <480*640>. [2]Saeed Rastegar, Reza Ghaderi,
Bounding box instruction detects matrices, Gholamreza Ardeshipr & Nima Asadi, " An
which contain the coordinates of upper left intelligent control system using an efficient
corner of the bounding box and specifies the License Plate Location and Recognition
width of the bounding box along each Approach", International Journal of Image
dimension. Processing (IJIP) Volume(3), Issue(5) 252,
And then for each connected element we 2009.
draw a rectangle as we can see in the code. [3]Wisam Al Faqheri and Syamsiah
Mashohor, "A Real-Time Malaysian
We did the same previous steps for a Automatic License Plate Recognition (M-
template matching image. ALPR)using Hybrid Fuzzy" ,IJCSNS
After that we create a function to separate International Journal of Computer Science
the charsets in the num_of_plate image and and Network Security, VOL.9 No.2,
the template matching image. February 2009.
[4]Satadal Saha1, Subhadip Basu, Mita
Conclusion: Nasipuri, Dipak Kumar Basu,” License Plate
We have implemented number plate Localization from Vehicle Images: An Edge
recognition. Our algorithm successfully Based Multistage Approach”, International
detects the number plate region from the Journal of Recent Trends in Engineering,
image which consists of vehicle number & Vol 1, No. 1, May 2009.
then character segmentation, recognition [5]Loumos, V.; Kayafas, E.,” License plate
.We have applied our algorithm on many recognition from still images and video
images and found that it successfully sequences: A survey” IEEE Transactions on
recognition. The project was designed Intelligent Transportation Systems,
keeping in mind the automation of the volume9, issue3, pages 377-391, September
number plate detection system for security 2008.
reason that could replace the current system

Volume 23, Issue 2, February - 2021 Page-369

Journal of University of Shanghai for Science and Technology ISSN: 1007-6735

[6]Ganapathy and W.L.D. Lui, "A

Malaysian Vehicle License Plate
Localization and Recognition System",
Journal of
Systemic,CyberneticsandInformatics,Vol.6,
No.1,2008.
[7]Roushdy M.,"Comparative Study of Edge
detection Algorithms Applying on the
Grayscale Noisy Image Using
Morphological filter", ICGST, International
Journal of Graphics, Vision, and Image
Processing GVIP, Vol.6, Issue4, pp. 17-23,
Dec. 2006.
[8]Chirag N. Paunwala, Suprava Patnaik, “A
Novel Multiple License Plate Extraction
Technique for Complex Background in
Indian Traffic Conditions”, In Proceedings
of International Journal of Image
Processing, vol.4, issue2, 2007
[9]D.Zheng, Y.Zhao, and J.Wang, “An
efficient method of License Plate location,”
Pattern Recognit. Lett. vol. 26, no. 15, pp.
2431– 2438, 2005.
[10]Feng Yang and Zheng Ma. “Vehicle
License Plate Location Based on
Histogramming and Mathematical
Morphology”, Automatic Identification
Advanced Technologies, pp: 89 –94,2005

Volume 23, Issue 2, February - 2021 Page-370