International Association of Scientific Innovation and Research (IASIR) ISSN (Print): 2279-0047

(An Association Unifying the Sciences, Engineering, and Applied Research) ISSN (Online): 2279-0055

International Journal of Emerging Technologies in Computational

and Applied Sciences (IJETCAS)
www.iasir.net

IMAGE PROCESSING BASED TOLL AUTOMATION USING ANPR

1
Kavya.S, 2Mutturaj Hubballi, 3Priyanka.B.N
1,2,3
Department of Electronics & Communication Engineering,
BMS Institute of Technology and Management,
6446 Doddaballapur Main Road, Avalahalli, Yelahanka,
Bangalore, Karnataka 560064, INDIA.

Abstract: This paper presents a review of Automatic Number Plate Recognition (ANPR) algorithm using image
processing in toll booths. On any toll booth the vehicle has to stop for paying the toll. We are trying to develop a
system that would pay the toll automatically and reduce the queue at the toll booths. In this system camera is
used for capturing the image of the vehicle number plate. The captured image would be converted into text
using ANPR algorithm. Toll booth passes the vehicle plate number to the central server which holds the
information about registered users. Server then uses this vehicle number to verify whether the user is registered
account holder or not. If the user is registered account holder then server transmits the information about
vehicle type, account balance to the toll booth. Based on the vehicle type, toll is deducted from the customer’s
account and then the toll gate is opened. Central server is updated with user’s account balance using internet
protocols by the toll booth the vehicle passes through. The user is updated about his account status via SMS.
Keywords: Automatic Number Plate Recognition, Optical Character Recognition, SMS, Image processing,
Server.

I. Introduction

Toll gates are usually considered as an inconvenience by travelers not only for the cost of toll, also for the delay
at toll booths. In order to ensure a steady flow of traffic, both staff and drivers require easy access to an efficient
communication system covering the specific requirements of toll gates.

II. Related works

“Automatised Toll Gate System Using Passive RFID and GSM Technology”[1]. This paper proposes a system
using RFID wherein the reader is placed in a strip which is laid beneath the lane, and the tag is placed in the
front side of the number plate. The object detection sensor which is placed on the side of the road detects the
approach of the incoming vehicle and intimates the stepper motor to raise the strip. Thus the reader rises to
ground level and reads the information in the tag and the transaction takes place through a centralized database.
“Electronic Toll Collection System Using Barcode Laser Technology”[2]. This paper proposes a barcode laser
system which uses tags (barcode) that are mounted on the number plate of vehicles, through which information
embedded on the barcode are read by barcode readers. The drawback of this system is it requires direct line of
sight and barcode scanner needs to be close to it.
“Image Processing Based Automatic Toll Booth in Indian Conditions”[3]. For the identification of the stolen
vehicles, the information of the vehicles is already stored on the central database. So captured number will be
sent to the server received at the toll. In case it passes through any other toll a silent alarm would buzz which
indicates the operator that the vehicle is stolen.
“Automatic Number Plate Recognition”[4] This paper presents two fastest algorithms used are edge finding
method and window filtering method for the better development of the number plate detection system. In this
method even after filtering is applied additional areas of high intensity appear beside number plate.
“Automatic License Plate Recognition (ALPR):A State-of-the-Art Review” [5] Presents a comprehensive
review of the state of the art techniques for ANPR(Automatic number plate recognition ).The algorithm
proposed is designed to recognize license plate of vehicle automatically based on video. since it is video based
it needs frame differencing or motion detection especially when the vehicles are at high speed.
This paper presents an automatic method of toll collection which overcomes the limitations of manual methods.
Captured image of the vehicle’s number plate is sent to a system with MATLAB software running in it, which
pre-processes the image and extracts the vehicles plate number. Extracted number is then sent to central server
to access the user information. Based on the type of vehicle toll amount is deducted from the user’s account.
Central server is updated about the users account every time a vehicle passes through a toll and the information
is accessed by the next toll booths later.TCP/IP protocols are used to accomplish the same. User is informed
about his account status through a SMS.

IJETCAS 15-351; © 2015, IJETCAS All Rights Reserved Page 140

Kavya et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 12(2), March-May 2015, pp. 140-143

III. Proposed System

A. Automatic Number Plate Recognition

ANPR can be used to store the images captured by the cameras as well as the text from the license plate.
Systems commonly use infrared lighting to allow the camera to take the picture at any time of the day. ANPR
technology tends to be region-specific, owing to plate variation from place to place. ANPR uses optical
character recognition (OCR) on images taken by cameras. Some license plate arrangements use variations in
font sizes and positioning ANPR systems must be able to cope with such differences in order to be truly
effective. More complicated systems can cope with international variants, though many programs are
individually tailored to each country. Some systems use infrared cameras to take a clearer image of the plates.
There are four primary algorithms that the software requires for identifying a license plate:
1. Plate localization – responsible for finding and isolating the plate on the picture.
2. Preprocessing – adjusts the brightness and contrast of the image.
3. Character segmentation – finds the individual characters on the plates.
4. Optical character recognition – The segmented characters are recognized by template matching.
B. Plate Localization
Captured image is converted to grey scale and smoothened to remove noise using median filter. Adaptive
histogramming is performed to enhance contrast of the image. Discrete wavelet transform is performed using
Haar filter to enhance the vertical edges. The intensity values are summed row by row and the row with the
maximum value is identified. The neighbouring rows above and below the row with maximum intensity within
an approximated number is extracted.
Figure 1 Cropped Image

C. Preprocessing
Cropped plate region is subjected to noise filtering and is further processed using structural element for
morphological operations to enhance contrast of the image. Edges are brightened using convolution function..
The image still has horizontal lines that are borders of name plate which is unwanted this can be removed by

IJETCAS 15-351; © 2015, IJETCAS All Rights Reserved Page 141

Kavya et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 12(2), March-May 2015, pp. 140-143

using structuring element ‘line’ with angle 00 to eliminate all horizontal lines in the image using imopen
function. Image thinning is used to remove very fine and small noises.
Figure 2 Preprocessed Image

D. Character Segmentation
In this step, the characters & digits of the plate are segmented and each is saved as a different image. The pre-
processed image is given as input to segmentation function. The objects in the image are identified by labeling
and plotting bounding boxes around them. Each object identified is taken and only those objects with number of
rows and columns within the predefined number range are considered and the corresponding characters are
extracted which gives the number plate.
Figure 3 Segmented Image

E. Optical Character Recognition

OCR is done using template matching. Images from A to Z and numbers from 1 to 9 are taken into different
variables and are pre-processed. All these variables are stored in the form of a cell in which each sub matrix
represents a letter. The same process is done for printed upper case, printed lower case, and printed numbers,
hand written upper case, hand written lower case and hand written numbers. All the model inputs are saved
under the same variable name. Every character extracted from the number plates are compared with every
template using correlation function, matching index is calculated and the one with maximum index value is
found. The letter corresponding to that value is taken as the number or alphabet, which is repeated for every
character obtaining entire registered number for further processing.

Figure 4 OCR Output

F. Database
The central database is the heart of the whole database system. This central database will be managed by a
central administrator. The customer has to be registered for this account to use this system. Database includes all
registered vehicles with details about owner, vehicle number and account balance. When the registered customer
passes through a specified toll booth then automatically the toll fee will be deducted from customer's account.
Central database is updated with this information at a same time, the customer will receive the information
through SMS to his mobile using GSM modem.

Figure 5 Indicating Amount Deduction

IJETCAS 15-351; © 2015, IJETCAS All Rights Reserved Page 142

Kavya et al., International Journal of Emerging Technologies in Computational and Applied Sciences, 12(2), March-May 2015, pp. 140-143

IV. Conclusion
Our system is a user friendly toll fee method which can save time and reduce traffic congestion at toll gates and
provide solution for users to reach their destination without wastage of time. It can be used to remove all
drawbacks with the current system such as time and human effort and it also doesn’t require any tag only
requires best quality camera and fixed font number plate. The main drawback of our paper is oriented number
plates and false detection of objects in segmentation step. The future research of ANPR should concentrate on
multi style plate recognition, oriented plates, high definition plate image processing, ambiguous-character
recognition.
Reference
[1]. Ward Nicholson, “License Plate Fonts of the Western World”, Available at http://
www.leewardpro.com/articles/licplatefonts/licplate-fonts-intro.html
[2]. Shan Du, Member, IEEE, Mahmoud Ibrahim, Mohamed Shehata, Senior Member, IEEE and Wael Badawy, Senior
Member,IEEE ,“Automatic License Plate Recognition (ALPR):State of the Art Review”, Available at “ieeexplore.ieee.org” Vol
23, 07 june 2012.
[3]. M. M. Shidore, S. P. Narot, “Number Plate Recognition for Indian Vehicles”, in IJCSNS, Vol 11 No.2, Feb. 2011
[4]. S Saha,S Basu, M Nasipuri, D K Basu,“License Plate Localization from Vehicle Images: An Edge Based Multi-stage Approach”,
in International Journal of Recent Trends in Engineering (Computer Science), Vol 1, No. 1, pp. 284-288,May2009.
[5]. Ankush Roy Debarshi, Patanjali Ghoshal, “Number Plate Recognition for Use in Different Countries Using an Improved
Segmentation”, Available at http://ieeexplore.ieee.org, Vol2, June 2012
[6]. S.Kranthi, K.Pranathi, A.Srisaila,” Automatic Number Plate Recognition”, available at http://ijict.org/, Vol 2, No 3 (July 2011).
[7]. Subhash Tatale,Akhil Khare,“Real time ANPR for vehicle identification using neural network”, ijaet, Vol. 1, Issue 4, pp. 262-
268, Sept 2011
[8]. “ANPR Manual - Tech liberty NZ”, Available: http://techliberty.org.nz/wp-content/uploads/ANPR-manual.pdf
[9]. “Janani.S.P, Meena.S.” Automatized toll gate system using passive RFID and GSM technology. Feb 10 th2012
[10]. “sanchit agarwal,shachi gupta,Nidhee sharma.”Electronic toll collection system using barcode laser technology, march- april
2014

Acknowledgements
We would like to thank Ms Ambika R, Associate Professor, Department of Electronics and Communication Engineering, BMSIT & MGMT
for providing constant encouragement, guidance and valuable suggestions without which this paper wouldn’t have been possible.

IJETCAS 15-351; © 2015, IJETCAS All Rights Reserved Page 143