See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/312529686

Vehicle Plate Number Detection and Recognition Using Improved Algorithm

Article · January 2014

CITATION READS
1 494

4 authors, including:

Michael Kisangiri Nerey H Mvungi

The Nelson Mandela African Institute of Science and Technology University of Dar es Salaam
15 PUBLICATIONS   29 CITATIONS    99 PUBLICATIONS   486 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

iGrid/smart grid View project

Antenna analysis View project

All content following this page was uploaded by Nerey H Mvungi on 19 January 2017.

The user has requested enhancement of the downloaded file.

Computer Engineering and Intelligent Systems www.iiste.org
ISSN 2222-1719 (Paper) ISSN 2222-2863 (Online)
Vol.5, No.10, 2014

Vehicle Plate Number Detection and Recognition Using

Improved Algorithm
Cosmo H. Munuo
School of Computational and Communication Science and Engineering, Nelson Mandela Institute of Science
and Technology,
P.O Box 447 Arusha, Tanzania munuoc@nm-aist.ac.tz

Dr Michael. Kisangiri
School of Computational and Communication Science and Engineering, Nelson Mandela African Institution of
Science and Technology,
P.O Box 447 Arusha, Tanzania, kisangiri.michael @nm-aist.ac.tz

Eng. Prof. Nerey H. Nvungi,

Principal of College of Information and Communication Technologies, University of Dar es Salaam,
P.O. Box 35194, Dar es Salaam, Tanzania, nhmvungi@udsm.ac.tz
Abstract
The growing Tanzanian population currently estimated to be 48 Million people and their use of vehicles as means of
transport has kept increasing making enforcing traffic rules and regulations among road users a major challenge. This calls
for a need to have an automated system that monitors the motorists with a pre-defined sense of intelligence. A Vehicle
Detection and Recognition Algorithm which can provide automated access to relevant information to a number plate from
information systems containing and managing databases on vehicle and their movements is required. This paper presents
work on developed algorithm that localizes plate area, extract and segment character, and finally recognizes and interprets
registration number from vehicle image. MATLAB R2012b Simulation software with Image Processing toolbox is employed.
HSV color space image, morphological and statistical analysis operations were integrated and employed to a vehicle image to
compute plate number area. In segmentation the properties like aspect ratio, extent, and area ratio were important
measurement parameters. Finally, the template matching database and statistical character extracted from car image was
correlated to recognize alphanumeric character to deduce car registration number.

Keywords— Character extraction, Detection algorithm, Recognition algorithm, Morphological matching, Template
matching

1 INTRODUCTION
TANZANIA population has double since 1988, with a currently estimated population of 48 Million people[1, 2]
whose use of vehicles as means of transportation and number of vehicles has largely increased and keeps
increasing. The vehicles usage includes; personal, public and/or business use. With this rapid increase, enforcing
traffic rules and regulations among road users becomes particularly difficult because of their habit to disregard
them. This calls for a need to have an automated system that monitors the roads users with a pre-defined sense of
intelligence. There is no such system implemented in Tanzania in public roads, hence an algorithm is proposed
that can automatically detect a vehicle plate number, isolate it from the rest of the car image, and recognize the
plate’s alphanumeric characters is required.
Since, Vehicle number plate is the only unique code for vehicle identification in any country database; it is an
important resource for accessing data of the owner and other relevant information from the country information
systems held at Tanzania Revenue Authority data bank. The United Republic of Tanzania has variety of number
plate’s registration standards a sample of which is shown in Fig. 1.
Therefore, the information from number plates can be used for laws enforcement, road toll data collection,
weighing bridges, police barriers, etc. It may also be used to facilitate other automated processes like automatic
opening/closing of gates, keep a time record on the entry or exit of vehicles, parking lot payment calculations
[3]. Therefore, by simply thinking number plate, one can imagine other applications such as traffic control,
journey time management, carjacking protection and border control because image is worth than thousands
word.

31
Computer Engineering and Intelligent Systems www.iiste.org
ISSN 2222-1719 (Paper) ISSN 2222-2863 (Online)
Vol.5, No.10, 2014

Fig. 1 Sample of Tanzanian Number Plate

Vehicle Number Plate Detection and Recognition (VNPDR) has a wide range of applications since the plate
number is the primary, most widely accepted, human readable, mandatory identifier of motor vehicles. VNPDR
provides automated access of the contents of the number plate form computer systems managing databases and
processing information of vehicle movements.
This paper is arranged as follows: section II is literature reviews, III experimental results and discussion and
VI conclusion.

2. LITERATURE REVIEWS
2.1 Introduction
From literature review a lot have been done and employed to improve the system performance and various
researches carried out in this area. Technically, the technology is a challenging research area which has been
enabled by innovation in computers and sophisticated high resolution infrared cameras. This made easier image
processing techniques applicable to analyzing and extracting important features for plate numbers detection and
recognition [4, 5]. The literature reviews looked at Image pre-processing, number plate detection and extraction,
character extraction and segmentation, and finally recognition and interpretation.
2.2 Image Acquisition and Pre-Processing
Image acquisition is performed by high end IR cameras, placed at strategic positions to avoid obstacles in
order to obtain accurate images. Image acquired by camera always reflect the camera settings; among many
include color and hue, saturation and value or brightness (HSV) whereby essentially an image can be in its
natural form or slightly altered. It was noted that, color images are complex in space and time.
Pre-processing aims at image enhancement and restoration. This process eliminates noise, highlight edges and
improve the overall quality of an image. In image pre-processing an image goes through among many
procedures grey-scaling, dilation, erosion, and filtering and edge enhancement [4, 6]. Converting natural color
images, to HSV color space, to grey scale and then to binary is important as it reduces time and space
complexities. This concept was express by Reshma [6, 7] and Kim et tal [7].
2.3 Number Plate Detection and Extraction
Number plate extraction is a process of localization of number plate region(s) from car image. Localization
select right feature that will provide best results in number plate recognition stage. This is a difficult task due to
the enormous variations of shapes, sizes, color of a car, illumination conditions, texture and orientation of
number plate in car images. Therefore, in order to have reliable number plate detection algorithms, in such
dynamic environments, several choices have to be considered.
Several features that could be deployed to extract a rectangular shape of number plate from a car image
included color feature, aspect ratio, and texture edge density and shape/size of ROI, but for better detection rate
the combination of features could provide more reliable solution. HSV color space and integral image properties
could also be employed to locate the coordinates and position of yellow number plate and non-yellow number
plates. From Liu et tal and Kim et tal [8, 9] HSV enables one to find out the four coordinates of a rectangular
shape containing English like symbols or texture from which X and Y coordinates, Width and Height could be
extracted.
The process of number plate area detection and extraction involved three stages of finding area of interest
(ROI), filtering out background and removals of false objects, and lastly computation of connected components
which provided best results of required region of interest. In the process of locating the regions of Interest the
HSV method was applied. MATLAB function was used to find the coordinates o (x, y, width, height) to provide
the regions of interest. The masked image was computed using MATLAB functions to obtain rows and columns
of our region of interest.
32
Computer Engineering and Intelligent Systems www.iiste.org
ISSN 2222-1719 (Paper) ISSN 2222-2863 (Online)
Vol.5, No.10, 2014

Summation of mask from HSV process and mask from Morphological process provided the best results for
computing connected components. The connected components were computed to provide the actual plate area
position or coordinates of the region of interest. The properties of ROI including bounding Box, area of object,
aspect ratio, extent, coordinates and perimeter of the object were considered to find the connected components.
By combining all the above properties it is possible to provide best results at later stage of segmentation and
recognition.
Finally, statistical analyses were applied to find the connected components to be used to locate number plate
accurately. This was performed to ensure that the global boundary was maintained, by providing upper and
lower limit of a ROI Left and Right limit of a ROI.
Dimensional properties of rectangle shape or objects using morphology methods of dilation, erosion and fill;
were also used by other researchers to increase correct plate localization [3]. In these methods the region that
passed through the qualifiers, and also passed as a number plate, its boundary box coordinates were used to crop
the original image and obtain the number plate only. The qualifier checked if the area of a region was not less
than specified threshold and length of the aspect ratio is within the limits (2:1 and 4.6:1 for our case).

Fig. 2 Yellow and white reflector with single and double rows number plates

The number plate detection and extraction returned four entries [X, Y, Width and Height] which could deal
with adjustment of ROI because the image may contain angle.
From the above methods, it was possible to eliminate the disturbances of the fake objects whose structure and
components were similar to the vehicle number plate, but did not match plate fixed color and coordinates. This
new idea of combining several properties enabled maximized recognition rate and efficiency. The idea assumed
that the white number plate area was situated somewhere between those black rows, hence by finding the largest
vertical arrays of white pixels, it was possible to detect the left and right edge of the number plates.
2.4 Character Extraction and Segmentation
From literature Character Extraction and Segmentation is the most difficult part of detection and recognition
process [10]. The factors that made the processes difficult included image noises, space mark, plate orientation,
illumination variance, character arrangements and plate frame. Traditional character segmentation includes those
based on extraction of connected components; segmentation based on thresholding, segmentation based on
region-growing and may others[11].
The image that passed pre-processing stage for number plate recognition that was localized in previous
sections was examined to extract individual components from rectangular area. Character extraction algorithm
consists of global and local searching. In global searching the objects borders classify which part is enclosed
with connected components while removing or erasing all smaller parts of the borders. The local searching
algorithms were required to label individual connected pixel (objects) white objects of the inputs image. The
black objects were conceived as background. The widest labeled black objects must be the borders [12].
The character extraction aspect of image analysis seeks to identify individual components or objects within
the image. Character extractions operate in two dimensions and therefore any derived description of objects
within the image should contain all relevant shape and size information originally contained within the stored
image. There are two types of rectangular Tanzanian number plates; one with two rows and other of a single row
see figure 2. Both types of number plates contain seven characters, but its rectangular shape has different aspect
ratio.
The character extraction and segmentation algorithm was required to determine the region boundaries of sub-
images within the image. The plate area had to be sectioned into several objects to provide the sub-regions of
interest. The aim of this function was to divide and extract the components which were most likely objects to be
characters (Letters and Numbers) and left the rest of other unwanted components[9, 10].
The parameters to be considered to extract the individual candidates from localized rectangular plate were

33
Computer Engineering and Intelligent Systems www.iiste.org
ISSN 2222-1719 (Paper) ISSN 2222-2863 (Online)
Vol.5, No.10, 2014

aspect ratio, extent, ratio of component area to the plate area, ratio component bounding area to the plate area of
the letters and numbers present in the in the plate. The measurement of above properties or parameters was
important factor of the recognition algorithms for subsequently manipulations of car image.
The choice of character to be derived from image influenced the effectiveness of interpretation stages in the
detection and recognition algorithms. Therefore, a set of character should be identified that could help to
uniquely indicate the key differences between the hopefully objects that would be uncounted constrained
environment of algorithms.

2.5 Character Recognition and interpretation

Character recognition was most important task in recognizing number plate. The recognition of character has
been a problem because there is a lot of possibility that character produced from previous stage may differ from
that in the database. The same character may differ in shape size and style, which could result in false character,
affect effectiveness and increase complexity of the system. This section looks at the algorithms which were used
to classify and then identify the individual characters. The recognitions were based on extracted features.
Recognition of characters had been a big task which received much attention [13-15] in the field of pattern
recognitions and image processing. Some previous researchers who contributed in recognition and
interpretations of characters proposed methods like Markov model, neural network approaches, statistical pattern
recognition, chain code, template matching and many others [16]. Each of these methods has advantages and
disadvantages dependent on the working environments.
This was the core of this research; the processor to recognize snapshot captured by camera, and return a text
representation of the detected image. This was the software aspect of the system which runs on standard PC
hardware and could be linked to other applications or databases. It first used a series of image manipulation
technique to detect, normalize and enhance the image and finally optical character recognition to extracts the
alphanumeric from the license plate.
Vehicle number plate detection and recognition algorithms work was based on assumptions made and it is
these assumptions that determined the weight of advantage and disadvantages in respect to algorithm’s level of
accuracy and robustness. Due to the varying characteristics of the license plate among countries/regions further
research is needed in this area [17-19]. Therefore, for better detection rate, accuracy and robust algorithms, the
combination of features and method can be used to provide more desirable results.

3. Experimental results and Discussion

The overall design processes of VNPDR followed the procedures based on MATLAB R2012b software. Some
function, script and command were required to perform VNPDR operations. MATLAB R2012b Simulation
software has Image Processing toolbox which was necessary when working with images. The whole process of
vehicle number plate detection and recognition could be achieved using a procedure as explained below.
3.1 Image Acquisition and Pre-Processing
After loading the image to the software the first stage in detection and recognition of number plates was to
perform image pre-processing that consisted of image enhancement and restoration processes for quality
improvement.
These were:
RGB to gray scale: conversion of natural color to gray scale image (rgb2grey()).
RGB to indexed image: representation of color image in a practical way. It stored image as two matrices. The
first matrix was instruction and second was color map command [X map] = rgb2ind() was applied.
HSV color space conversion: conversion of Natural color to HSV. RGB image was not preferred in detection
and recognition process because it was not stable when light conditions varied (hsv=rgb2hs ()).
Filtering: The media filter to eliminate the unwanted noise from the acquired image command medfilt2 (). The
fig.3 shows the procedure applied for pre-processing

3.1.1 HSV color space

Natural color or original RGB was not used in localization of plate number; instead HSV was preferred
because it was stable when light conditions varied. Therefore, it was easier to detect and recognize number plate
with varying color or for similar parts of color values of a plate region.
In MATLAB R2012b Function rgb2hsv () was used for color conversion image between the RGB and Hue,
Saturation, Value (HSV) color space. The HSV color space was often used for picking colors (eg paint or ink)
from color wheel or palette because it corresponded better to how people experienced color than RGB color

34
Computer Engineering and Intelligent Systems www.iiste.org
ISSN 2222-1719 (Paper) ISSN 2222-2863 (Online)
Vol.5, No.10, 2014

space.

Fig. 3 Flow chart to show image pre-processing

As shown in Fig. 4; Hue, Saturation, and Value or brightness helped to indicate exactly location of plate area
in the car image. When working on pre-processed image, it was important to put emphasis on for quality
enhancement and restorations.

b
Fig. 4 (a) show original images and (b) show HSV image.

3.2 Number Plate Area Detection and Extraction

The process of number plate area detection and extraction involved three stages of finding area of interest
(ROI), filtering out background and removals of false objects, and lastly computing of connected components
which provided best results of required region of interest see Fig. 5.

Fig. 5 Sequential flow chart to compute ROI

The masked image was computed using function (dim (1), dim (2)), where dim (1) computed rows and dim
(2) computed columns. MATLAB 2012b experiment simulated to test loop to obtain the mask that consisted of
regions of interest (ROI) with seven objects as shown in fig.6. Algorithm flow chart indicated in fig. 5 and fig. 7
was applied to compute ROI. Thereafter, the coordinates and position of number pates were located. The
algorithm in fig.8 determined the four coordinates of the rectangular shape which contained English like
alphanumeric character that was extracted as [X, Y, Width and Height]. Image shown in fig.6 indicate exactly
35
Computer Engineering and Intelligent Systems www.iiste.org
ISSN 2222-1719 (Paper) ISSN 2222-2863 (Online)
Vol.5, No.10, 2014

the plate number area. The HSV color space image, morphological and statistical analysis operations were
integrated to compute plate number area which was region of interest (ROI).

b
Fig. 6 (a) & (b) number plate area for single and two rows

From Pre-processing HSV image

Decrease Value Plate area

Compute dims: width(x) height(y) & mask(x,y)==1

Morphological Operation: Dilation, Erosion, fill

Connected components: max area, aspect ratio,

perimeter

Mask=1, Character=7

Output ROI with coordinates [X Y Width

Fig.7 shows flow chart output ROI coordinates

36
Computer Engineering and Intelligent Systems www.iiste.org
ISSN 2222-1719 (Paper) ISSN 2222-2863 (Online)
Vol.5, No.10, 2014

3.3 Character extraction and Segmentation

3.3.1 Experimental Analysis
The experimental analysis, cropped image of public or private vehicle which contained 7 characters, arranged
in single row or double rows was fed to segmentation stage. Character extraction algorithm involved global and
local boundaries searching. The character segmentation algorithm (s) was processed repeatedly while reducing
the threshold value until seven (7) connected components (objects) were found see fig. 8.

Initiate accumulator
Input ROI

N
Threshold & Test for 2 rows
filtering Y

Cut rows: upper & lower sections

Compute: upper& lower character Single Row

Accumulate: upper + lower character

N Character=7

Y Segmented Character

Fig. 8 Flow chart which returns segmented character

The design of algorithm to recognize plate that started with single letter followed by 3 numerical and
completed with recognition of 3 letters. An algorithm could test for double rows or a single row. If the output
was for double rows, the number plate was cut into upper and lower sections to maintain integrity of registration
number.
The criterion used to divide the plate into two parts was to ensure that, the plate did not cut the characters.
The row with minimum pixels was found; in the middle of two rows there was minimum whitish or reversed
black pixel which has be used to achieve this purpose. That means result of upper and lower rows would sum up
to form 7 characters, whereby the upper row contained 4 characters and lower rows contained 3 characters. The
output from single row also produced 7 characters at global boundary, that were identified from left to right side
of number plate whereby no cut action was applied.

3.3.2 Results
The algorithm detected either a single or a double rows number plate format using aspect ratio techniques.
Double rows format was define with aspect ratio less than 2.2. The threshold value was applied to reduced
components into separated characters. At threshold, all unwanted sub-objects were filtered out from the objects
of interest which was seven alphanumeric characters. The output of this experiment which has JPEG format and
42x24 pixels was fed to recognition stage for interpretations.
37
Computer Engineering and Intelligent Systems www.iiste.org
ISSN 2222-1719 (Paper) ISSN 2222-2863 (Online)
Vol.5, No.10, 2014

3.4 Character Recognition and Interpretations

3.4.1 Experimental analysis
The character recognition algorithm developed was demonstrated using MATLAB R2012b. The simple
matrix database template of each letter and number made of 18 fonts was created alone with algorithm. The
database contained letters A to Z and numbers 0 to 9 English like that are used in Tanzanian numbering system.
Before matching processes, the acquired image from source were processed and read as a JPEG. Therefore, in
recognition stage the extracted character from segmented image was differentiated from background.
Template matching
For experimental purpose, the template matching database and character extracted from car image was
correlated to recognize letters and numbers. For this case an extracted character were correlated with template
database which are made of 0 to 9 numbers and A to Z English letters. The size of each character image was
normalized according to the size of characters stored in the database see fig. 9. Here each character was
fitted to 42 x 24 was correlated for maximum match.

Fig.9 template database

3.4.1 Result
The developed algorithm recognized the characters and presented them as text. The method used was to
compare the extracted characters with the database called templates. The recognition was done by comparing
each extracted character to all the templates in the database and the one with maximum resemblance was picked.
Input extracted char Initiate count

Char<=7 Output reg. no. (7 character)

Noo

Input letter/ number

New char
Template Database

Add letter or number Temp=match char

N
Letter/ number

Fig. 10 Flow chart output the registration number

Each extracted character was correlated to its maximum corresponding pixel in the template. The character
was validated when the maximum match was achieved. In the validation process, if the seven characters
present in the number plate and characters present in the character database matched then validation was
successful and authorization was given to that particular vehicle for security and automated applications.
Fig.11 and fig.12 shows recognized registration number using developed algorithm as implemented in MATLAB

38
Computer Engineering and Intelligent Systems www.iiste.org
ISSN 2222-1719 (Paper) ISSN 2222-2863 (Online)
Vol.5, No.10, 2014

R2012b. the registration number is located on top of image.

(a) (b) (c) (a)

Fig. 11 (a), (b), (c), and (d) recognized registration number on top of image.

Fig. 12 Recognized registration number.

4. CONCLUSIONS
The concept introduced of using multiple factors to improve reliability of number plate recognition schemes
has been successfully tested for both single row and double number plates used in Tanzania. The process of
taking the image and processing timely. It has been shown that number plate detection algorithms require an
effective sets of hardware and software components to be implemented. Tests made showed that set of
processed input image helps to recognizes number plate character from car image background independent of
location, size or orientation of the objects, so that the variation in number plate position and orientation can be
accommodated.
The output of this research can help to cut down human resource cost and increase overall efficiency of
transportation and traffic monitoring sector. It can also support fighting against crimes and offenses like terrorist
and blacklisted vehicles. In summary in can be applied in a number of enforcement applications.

ACKNOWLEDGEMENT
The authors greatly acknowledge the financial support provided by Tanzanian Government and Nelson
Mandela African Institute of Science and Technology (NM-AIST) for imparting knowledge and facilitating
research work.

REFERENCES

[1] J. E. Kocher, "Tanzania: population projections and primary education rural health and water supply goals," Development
Discussion Paper, 1976.
[2] M. Schreiner, "A Simple Poverty Scorecard for Tanzania," ed: Progress out of Poverty, Washington, DC Available from:
http://progressoutofpoverty. org/tanzania (enrolment and login required)[accessed 4 June 2012], 2011.
[3] W. Weihua, "License Plate Recognition Algorithm Based on Radial Basis Function Neural Networks," in Intelligent Ubiquitous
Computing and Education, 2009 International Symposium on, 2009, pp. 38-41.
[4] S. G. Patel, "Vehicle license plate recognition using morphology and neural network," Int. J. Cybernetics Inform, 2013.
[5] I. Bulugu, "Algorithm for License Plate Localization and Recognition for Tanzania Car Plate Numbers," International Journal,
2013.
[6] P. Reshma, "Noise Removal and Blob Detection Approach for Number Plate Identification," International Journal of Computer
Applications, vol. 47, pp. 13-16, 2012.
[7] J. Kim, Y. Han, and H. Hahn, "License Plate Detection Using Topology of Characters and Outer Contour," in Future Generation
Communication and Networking Symposia, 2008. FGCNS'08. Second International Conference on, 2008, pp. 171-174.
[8] F. Wang, L. Man, B. Wang, Y. Xiao, W. Pan, and X. Lu, "Fuzzy-based algorithm for color recognition of license plates," Pattern
Recognition Letters, vol. 29, pp. 1007-1020, 2008.
[9] L. Liu, H. Yu, K. Cai, and J. Wang, "License plate recognition using topology structure features," in Computing, Control and
Industrial Engineering (CCIE), 2011 IEEE 2nd International Conference on, 2011, pp. 251-254.

39
Computer Engineering and Intelligent Systems www.iiste.org
ISSN 2222-1719 (Paper) ISSN 2222-2863 (Online)
Vol.5, No.10, 2014

[10] J. Kim, Y. Han, and H. Hahn, "Character segmentation method for a License plate with topological transform," World Academy
of Science, Engineering and Technology, pp. 39-42, 2009.
[11] S. Kranthi, K. Pranathi, and A. Srisaila, "Automatic number plate recognition," Int. J. Adv. Tech, vol. 2, pp. 408-422, 2011.
[12] M.-P. Dubuisson Jolly, S. Lakshmanan, and A. K. Jain, "Vehicle segmentation and classification using deformable templates,"
Pattern Analysis and Machine Intelligence, IEEE Transactions on, vol. 18, pp. 293-308, 1996.
[13] N. K. Ibrahim, E. Kasmuri, N. A. Jalil, M. A. Norasikin, S. Salam, and M. R. M. Nawawi, "License Plate Recognition (LPR): A
Review with Experiments for Malaysia Case Study," arXiv preprint arXiv:1401.5559, 2014.
[14] G. Guangying, B. Xinzong, and G. Jing, "Study on automatic detection and recognition algorithms for vehicles and license plates
using LS-SVM," in Intelligent Control and Automation, 2008. WCICA 2008. 7th World Congress on, 2008, pp. 3760-3765.
[15] A. Akoum, B. Daya, and P. Chauvet, "TWO NEURAL NETWORKS FOR LICENSE NUMBER PLATES RECOGNITION,"
Journal of Theoretical & Applied Information Technology, vol. 12, 2010.
[16] Y. Wen, Y. Lu, J. Yan, Z. Zhou, K. M. von Deneen, and P. Shi, "An algorithm for license plate recognition applied to intelligent
transportation system," Intelligent Transportation Systems, IEEE Transactions on, vol. 12, pp. 830-845, 2011.
[17] K. Romić, I. Galić, and A. Baumgartner, "CHARACTER RECOGNITION BASED ON REGION PIXEL CONCENTRATION
FOR LICENSE PLATE IDENTIFICATION," Tehnicki vjesnik/Technical Gazette, vol. 19, 2012.
[18] S. Saha, S. Basu, M. Nasipuri, and D. K. Basu, "License Plate localization from vehicle images: An edge based multi-stage
approach," International Journal on Recent Trends in Engineering (Computer Science), vol. 1, pp. 284-288, 2009.
[19] L. Xiaoping, L. Xiaoxing, W. Shuaizong, Z. Lin, L. Yinxiang, and D. Hongjian, "Research on the recognition algorithm of the
license plate character based on the multi-resolution template matching," in New Trends in Information Science and Service Science (NISS),
2010 4th International Conference on, 2010, pp. 341-344.

Author Profile

Cosmo H. Munuo, received PGD in Electronics and IT from University of Dar es Salaam (UDSM),
Advanced Diploma in Electronics and Telecommunications from Dar es Salaam Institute of Technology
(D.I.T) and Full Technician Certificate (F.T.C) in Electrical Engineering from Technical College Arusha.
Presently, a Master degree student of Information and communication Science and Engineering specialized
in Electronics and Telecommunications at Nelson Mandela Institution of Science and Technology (NM-
AIST, Arusha), Tanzania. Currently is a Member of Engineer Registration Board of Tanzania. Sponsored
and work at Tanzanian Government as Telecommunication Engineer. His research interest includes digital
Image processing, wireless communication and electromagnetic interference and control.

40
 The IISTE is a pioneer in the Open-Access hosting service and academic event
management. The aim of the firm is Accelerating Global Knowledge Sharing.

More information about the firm can be found on the homepage:

http://www.iiste.org

CALL FOR JOURNAL PAPERS

There are more than 30 peer-reviewed academic journals hosted under the hosting
platform.

Prospective authors of journals can find the submission instruction on the

following page: http://www.iiste.org/journals/ All the journals articles are available
online to the readers all over the world without financial, legal, or technical barriers
other than those inseparable from gaining access to the internet itself. Paper version
of the journals is also available upon request of readers and authors.

MORE RESOURCES

Book publication information: http://www.iiste.org/book/

IISTE Knowledge Sharing Partners

EBSCO, Index Copernicus, Ulrich's Periodicals Directory, JournalTOCS, PKP Open

Archives Harvester, Bielefeld Academic Search Engine, Elektronische
Zeitschriftenbibliothek EZB, Open J-Gate, OCLC WorldCat, Universe Digtial
Library , NewJour, Google Scholar

View publication stats