International Journal of Scientific & Engineering Research, Volume 7, Issue 5, May-2016 240

ISSN 2229-5518

Modified Number Plate Localization

Algorithm and its Implementation Using
FPGA
Rohini B. Watte, Prof. Dr. Janardan Chitode

Abstract-This paper represents implementation of number plate localization for automatic toll collection using
FPGA. Number plate localization plays a key role in various applications. Number of techniques has been
proposed for number plate localization. Number plate recognition is the process of extraction of any vehicle
number plate information from an image which is taken from a video. Since it is easier and faster than the normal
token based ticket system, it has all the capability to interchange the usual system. Hence it avoids the user to
wait in long queue on the counter of toll. It saves the valuable time of the users. It accustomed pay the bill
automatically and does the opening and closing of the gate automatically. A detection device is placed at the gate
which reads the information from the coming moving vehicle and compares it with stored database. After that it
gives permission to access consequently to move through the gate. This information is used to print a daily or
monthly basis bill for toll collection from the vehicles. This system has low complexity and it takes less time to
localise and recognise the characters on number plate. This system helps to assist RTO, local department to
trace the vehicle, if it was actually stolen or used for any illegal activities. In this paper, FPGA implementation is
done using VHDL language.

Index Terms: Number plate localisation, character segmentation, character recognition, binarization.

IJSER
1 INTRODUCTION applicable for FPGA implementation using
VHDL. A MATLAB implementation used in
Number plate localisation systems are used for
this algorithm is as proof of idea prior to
various applications such as track the vehicle if it
hardware implementation.
is stolen, unbilled parking heaps, access control,
and toll collection and identify the vehicle which
is used in illegal activities.
Number plate recognition techniques vary 2 PROPOSED TECHNIQUE
from application to application. The cameras
2.1 Number plate recognition
mounted on toll or anywhere, capture an image
as they move. This image could have contained In many of the plate detection algorithms has
different number plates or no number plate. more than class which supports different
After detecting the plate from an image, number techniques. To identify the variety of plates.
plate has different characteristics such as various Below points should be considered.
sizes and shapes of characters, orientations & 1) Size of plate: Plate can be of any size
positions. depending on the choice of the user.
In first stage of process position of number
plate present on vehicle is determined by 2) Location of plate on vehicle: Plate can be
adjusting the resolution, height & width of the placed at any place which varies on the type of
image using MATLAB. Character segmentation vehicle.
is used to extract number plate. There are
3) Background of the plate: colour of the plate
number of segmentation techniques, from which
can vary. In India 3 types of coloured plates are
many of the techniques uses binarization. After
used.
that character recognition is done by comparing
the types of characters which are various in sizes
A) White Background: white background with
and shapes stored in database.
black letters used only for private use. It can’t be
In this paper, the algorithm proposed is
used for commercial purpose.
————————————————
B) Black: These vehicles have black plate with
• Rohini B. Watte is currently pursuing master degree in Electronics yellow letters can be used for any commercial
(VLSI) in Bharati Vidyapeeth Deemed University College Of Engineering,
Pune, India. E-mail: [email protected] use without limitations of driver having to have
• Dr. Janardan Chitode is professor in Bharati Vidyapeeth Deemed
commercial driving license.
University College Of Engineering, Pune, India. E-mail:
[email protected]

IJSER © 2016
http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 7, Issue 5, May-2016 241
ISSN 2229-5518

C) Yellow: Yellow plate with black letters used Exact location of plate can be determined by size
for any commercial purpose. Driver should have of plate. To achieve these values of number of
commercial driving license. bands of rows and columns has to be calculated
with plate’s properties such as height & width.
There are number of techniques used for To do so [r c p]=size() function is used. So that
segmenting the image. In most of the techniques we can find exact location of the plate.
binarization is utilized. In following section plate
recognition is explained. Binarization is done by converting a gray image
into black & white. This by choosing certain
threshold value to classify some pixels as black &
Input Licence CC Splitted
Image Plate Analysis Characters
white. After binarization to fill complete black &
white colour in an image bwfill function is used.

Templates Feature
Matching After filling black & white colour into an image
Extraction
denoising process is carried out using median
Recognized characters filter. Median filtering is applied for both rows &
columns values so that we can get denoised
image.
Recognition Comparison with
Result database Character segmentation includes MATLAB
function named as regionprops() which
Fig 1 Number Plate Recognition measures set of properties of image region
boundingbox is used. At the end after labelling
Above system involves following steps:
the components, region will be extracted from
input image.

IJSER
1)Input Image 2) Binarized gray scale image
3)Locating plate 4)Black and white filling in
After segmenting characters & numbers, the final
image 5) Denoising 6) Character segmentation
process is recognition of characters. This can be
7)Character recognition.
done by matching or comparing the extracted
characters or elements with stored database
First input image is converted from RGB to gray
numbers and finally exact characters are carried
using rgb2gray() function and gray value is
out.
calculated using gray(i,j)=0.59*R(i,j) + 0.30*G(i,j)
+0.11*B(i,j).
The proposed system makes sure that the load of
traffic at toll is handy & security is good. The toll
tax is collected depend. On the carried load of
the vehicle. Readings taken from the information
of vehicle’s plate recognition system, computer
system compares the information with database
& gives permission to access by opening or
closing the gate. This information is used to print
daily or monthly bill for toll from vehicles.

Fig2 Input Images

Fig 3.Database Image

Fig.4 Block Diagram for Toll Gate System

IJSER © 2016
http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 7, Issue 5, May-2016 242
ISSN 2229-5518

2.2 Pressure Sensor

To count number of vehicles passed by the gate,
2.4 Wireless Communication Module using
pressure sensing element is used. These counted
RF transmitter & receiver.
numbers are displayed on screen. This pressure
sensing element also measures the weight of If any sensor detects any illegal gas or any
vehicles, depend on weight measured, kind of dangerous gas illegally, the buzzer gets
predefined amount i.e. tax will be taken. This is activated & alert signal is sent to nearby police
also displayed on screen. For this purpose we are station through RF transmitter through antenna
using piezo pressure sensor. & then this police station will receive this alert
signal via RF receiver through antenna. This is
done by RF communication strategy as shown in
fig.

Fig5 Piezo electric sensor

The amount to be paid is deducted from user’s Fig7 RF communication block diagram
checking account. These sensors are used to
measure the electric effect caused by applying RF transmitter & buzzer is connected to FPGA
force to piezoelectric material. It converts energy which takes corresponding decisions. Most of RF

IJSER
into mechanical & electrical energy forms. These communication modules are TTL compatible &
sensors are used to measure pressure, input can be given in TTL level logic. This TTL
acceleration, strain or force. input is converted into serial data input & read
by RF transmitter transmit through antenna at
2.3 Gas Sensor receiver end, it receives modulated signal &
Gas sensing element is used when a vehicle performs processing, filtering, demodulation &
carries any flammable gas that shouldn’t be give serial data out. This output is converted
carried. If flammable gas is detected, RF back to TTL logic level which is same as input
transmitter is used to alert nearby police station signal. After that action will be taken
associated. Buzzer is activated to alert nearby accordingly.
area. After ringing the buzzer, motor 1 is used to
shut down the gate & simultaneously motor 2
opens up spikes to puncture that vehicle. 3. RESULTS

Fig6 Gas Sensor (type-MQ2)

To sense flammable gas or any kind of smoke

present at the vehicle. Gas sensor MQ2 is used. It
is analog type of sensor. This sensor detects
pressure of smoke & sends an analog image to
FPGA. It is also used as gas leakage detecting
device. It detects H2, LPG , Alcohol, smoke or
Fig8 Input image is converted into Gray & Binary
propane. It is high sensitive & fast so smoke
Image
detection is done as soon as possible.

IJSER © 2016
http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 7, Issue 5, May-2016 243
ISSN 2229-5518

Fig12 After Matching

Fig 9 Denoised Number Plate

4 CONCLUSION
Main aim of this system is to recognize the
images of number plates. This system is
implemented to provide exact identity of
vehicles, captured images from camera. This

IJSER
system has been concentrated to detect lightning
variations, dim headlights, obscuration’s of
vehicles. This system also helps in collecting toll
tax of the user. The next step involves increasing
protection from illegal activities happening
nowadays on road every day.

5 FUTURE WORK
In future, this system will be concerned about to
Fig10 Before Segmentation Number Plate
focusing on different styles of plate recognition.
Video based high definition number plate image
process, multi-plates process. The important
points were planned to detect different varieties
of alphanumeric characters, rotational angle &
character formats.

REFERENCES
[1] XiaojunZhai, FaycalBensaali, Klaus McDonald-Maier,
“Automatic number plate recognition on FPGA”,
School of Computer Science & Electronic
Engineering,University of Essex, Colchester,
UK{xzhai, kdm}@essex.ac.uk, Department of
Electrical Engineering College of Engineering,Qatar
University, Doha, Qatar [email protected], 978-1-
4799-2452-3/13 IEEE 2013
Fig11 After Segmentation Number Plate
[2] XiaojunZhai• FaycalBensaali,” Improved number plate
character segmentation algorithm and its efficient
FPGA implementation”,Springer-Verlag 2012.

[3] Shyang-Lih Chang, Li-Shien Chen, Yun-Chung Chung,

Sei-Wan Chen,”Automatic license plate recognition”,
senior member, IEEE, IEEE transactions on Intelligent
transportation systems,vol.5.no.1, March 2004.

IJSER © 2016
http://www.ijser.org
International Journal of Scientific & Engineering Research, Volume 7, Issue 5, May-2016 244
ISSN 2229-5518

[4]XiaojunZhai, School of Computer Science & Electronic

Engineering, University of Essex, Colchester, UK,
[email protected], FaycalBensaali, Department of
Electrical Engineering, Qatar University, Doha,
Qatar, [email protected],” Standard definition
ANPR system on FPGA and an approach to extend it
to HD,978-1-4799-0724-3/13 IEEE GCC Conference
and exhibition, November 17-20, Doha, Qatar,2013.

[5] SarbjitKaur, Research Schloar ,SukhvirKaur, Assistant

Professor, CT Group of Institutions, Jalandhar,” An
Efficient Approach for Number Plate Extraction from
Vehicles Image under Image Processing, (IJCSIT)
International Journal of Computer Science and
Information Technologies, Vol. 5 (3) , 2014, 2954-2959.

[6]Amit Ashok Goregaonkar, TejasJnaneshGhalsasi,

RohitNeeleshSabnis, Fr. Conceicao Rodrigues
College of Engineering, Mumbai, Maharashtra, India,
”Automatic number plate localization and optical
character recognition”,The International Journal Of
Science &Technoledge (ISSN 2321 – 919X), Vol 3
Issue 3 March, 2015.

IJSER
[7] Dhiraj Y. GaikwadPramod B. Borole, Electrical
Department, V.J.T.I. MUMBAI, “A review paper on
Automatic number plate recognition (ANPR)
system”,International Journal of Innovative Research
in Advanced Engineering (IJIRAE) Volume 1 Issue 1
(April 2014).

[8] Dinesh Bhardwaj, Sunil Mahajan, Department of

Computer Science &Information Technology, SSM
College, Dinanagar, Punjab, India, “Review Paper on
Automated Number Plate Recognition Techniques”,
International Journal of Emerging Research in
Management &Technology ISSN: 2278-9359 (Volume-4,
Issue-5) May 2015

IJSER © 2016
http://www.ijser.org