Iaetsd an efficient way of detecting a numbers in car

An Efficient Way of Detecting a Numbers in Car
License Plate Using Genetic Algorithms
Abstract - To detect the numbers and characters inside the
license plate using image processing and genetic algorithm
(GA). For this Number plate detection many algorithms are
used. But in my project mainly focusing on the genetic
algorithm for provide perfect accuracy compare to any other
systems. This paper describes a detection method in which
the vehicle plate image is captured by the cameras and the
image is processed to get the plate’s numbers and
characters. The system is implemented using MATLAB and
various images are processed with to verify the distinction
of the proposed system.
Index Terms - Genetic algorithm (GA), Image processing,
License plate (LP), Number plate localization, Perfect
accuracy.
I.INTRODUCTION
Nowadays number of automobiles grows
quickly, the traffic problems arise as well, for example car
robbery, over speeding and moving on the red light. To
avoid these problems an efficient real time working
vehicle identification system is needed. Most usually
suitable technique is license plate (LP) detection based on
image processing by capturing license plates using
cameras. All the implemented techniques can be
classified according to the selected features. Color
information based systems have been built to detect
specific plates having fixed colors. Shape- based
techniques were developed to detect the plate based on its
rectangular shape. Edge-based techniques were also
implemented to detect the plate based on the high density
of vertical edges inside it. GAs has been used
infrequently because of their large computational needs.
Variety of research has been tried at different levels
under some constraints to minimize the search space of
genetic algorithms (GAs). Researchers in based their GA
on pixel color features to segment the image depending on
stable colors
followed by shape dependent policy to identify the plate’s
area. In, GA was used to search for the best fixed
rectangular area having the same texture features. GA
was used in to identify the LP symbols not to detect the
LP.
Detecting license character and at the same
time differentiating it from similar patterns based on the
geometrical relationship between the symbols
constituting the license numbers are selected approach
in this research. Consequently, a new approach genetic
algorithm is initiate in this paper that detects LP symbols
without using any information linked with the plate’s
external shape or interior colors to allow for the detection
of the license numbers in case of shape or color
distortion either physically or due to capturing
conditions. Further processes are explained in the next
sections.
II.PROPOSED TECHNIQUE
The proposed system is comprised of two
phases: image processing phase and GA phase. Each
phase is composed of many steps. The Fig. 1 depicts the
various image processing steps that finally produce
image objects to the GA portion. GA selects the best LP
symbol locations depending on the input geometric
relationship matrix (GRM).
III.IMAGE PROCESSING PHASE
In this phase, an input color image is used to a
sequence of processes to extract the relevant 2-D objects
that may represent the symbols. It has different stages, as
depicted in Fig. 1.
A. Color image to Grayscale conversion
he input image is used as a color image to
bring other information relevant to the concerned
vehicle. Color (RGB) to grayscale (gs) conversion is
S.Sudha M.E.,
Assistant Professor
Department of Computer Science and Engineering
Sree sowdambika College of Engineering Aruppukottai,
Tamilnadu St, India
sudhajiin@yahoo.co.in
C.Subha
M.E Computer Science
Sree Sowdambika College Of Engineering
Aruppukottai Tamilnadu St, India
mirthula31@gmail.com
Proceedings of International Conference on Developments in Engineering Research
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT
54

Fig. 1. Overall system Flowchart for localization of LP symbols.
performed using the standard NTSC method by removing the
hue and saturation information while holding the luminance
as follows:
gs=0.299*R+0.587*G+0.114*B (1)
Fig. 2. Converted grayscale image.
B. Grayscale to Binary Using Dynamic Adaptive Threshold
Converting the input image into a binary image is
one of the most important stages in localizing LPs to
overcome the illumination problems. In my system, a local
adaptive threshold technique has been implemented to
determine the threshold at each pixel depending on the
average gray level. This process as shown in Fig.3
C. Morphological Operations
Morphological operations, like dilation and
erosion, are important processes needed for pattern
recognition systems to eliminate noisy object.
Fig. 3. (a) Converted binary image for image in Fig. 2, using Otsu’s
method. (b) Car image with variable illumination. (c) Output when using
Otsu’s method for image in (b). (d) Output when applying local adaptive
threshold method for same image in (b).
In LP detection, closing and opening operations are applied
to fill noisy holes and remove objects.
Dilation:
This is the b asic operators in the part of
morphology. It is usually applied to binary image, but there
are versions run on grayscale image. the basic effect of the
operator on a binary image is to progressively extend the
boundaries of regions of foreground pixels (ie, white pixels).
Applications of dilation for bridging gaps in an image. It can
remove unwanted information. Opening of an image is
erosion followed by a dilation using the same structuring
element. Shown in fig 4.
Erosion:
This is also very important operator for the
morphological operation. The basic effect of the operator on
a binary image is to erode away the boundaries of regions of
foreground pixels (ie, white pixels). Shown in fig 5.
Fig 4. Effect of dilation using 3x3 square structuring element
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
55

Fig 5. Effect of erosion using a 3×3 square structuring element Strip away a
layer of pixels from an object, shrinking it in the process.
D. Connected Component Analysis
CCA is one of the technique in image processing that
scans an image and groups pixels in components depends on
pixel connectivity. The result of this stage is an array of N
objects.
E. Size Filtering
The output of the CCA stage are filtered on the
basis of their widths Wobj and heights Hobj lie between their
respective thresholds as follows:
Wmin ≤ Wobj ≤ Wmax and Hmin ≤ Hobj ≤ Hmax (2)
Hmin and Wmin are the value below which a symbol
cannot be recognized (for example 8 pixels) and Wmax can be
set to the image width divided by the number of symbols.
Hmax is estimated as Wmax divided by the aspect ratio of the
used font. The result of this stage is an array of M objects.
The output of this stage is given in Fig. 6.
Fig. 6. M objects (64) output after size filtering of N objects in Fig. 5(2).
IV.GENETIC ALGORITHM
In this phase M objects are given to the input. This
phase is used to resolve the 2D compound object detection
problem. It contains many steps.
A. Chromosome Encoding
In chromosome encoding an integer encoding scheme
is selected and each gene assigned to an integer. Seven genes
are forming a chromosome as shown in fig. 7. An output is
extracted as a M objects.
Fig 7. Chromosome of seven genes for representation of Saudi license plate.
B. Fitness Function
Simple function of the fitness measure is used by
some genetic algorithms to select individuals. In this proposed
system fitness is used as the inverse of the estimated objective
distance between the prototype chromosome and the current
chromosome.
C. Selection Method
In this selection method, the stochastic universal
sampling (SUS) method each individual is formed to a
continuous segment of a line. Depending on the percentage of
individuals to be selected by a number of pointers over the
line.
D. Mutation Operators
This mutation method is used to remove unfit
members in genetic iterations. It can eliminate some features
of genetic material. To maintain the mating pool variety by
Gas ensures that the new parts of the search space. They are
two kinds of mutation operators.
1) Substitution Operator
2) Swap Operator
E. Crossover Operator
In genetic algorithm crossover operator is used to
produce new chromosome (offspring) by groups two
chromosomes (parents). This new chromosome is better than
the both parents if it takes the best characteristics from each of
the parents. In my project, the two parents chromosomes are
combined into the array Carray as shown in fig.8. In my
project USPS crossover operator is used.
F. Replacement Strategy
A lot of alternate strategies are used to replacing only a
portion of the population between generations. The most
frequent strategy is to probabilistically replace the unfit
individuals in the earlier generation. In elitist strategy the
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
56

greatest fit individuals of the previous generation are
appended to the recent population. In my proposed
system, the best 10% of the parents are selected and
appended to the offspring(90%) to produce the new
generation (100%).
Fig. 8. Proposed crossover operator steps.
V.CONCLUSION
In this paper describes the localization of license plate in a
efficient manner. For this purpose i used genetic
algorithm (GA). The license plate contain many unwanted
details. These are first remove by the image processing
phase and then localized by the genetic algorithm phase.
The results were encouraging and a new approach for
solving the LP detection problem relying only on the
geometrical layout of the LP symbols. Also, a flexible
system was introduced that can be simply adapted for
any LP layout by constructing its GRM matrix. The
proposed system possessed high immunity to changes
in illumination either temporarily or spatially. A high
percentage success rate was achieved with the aid of
the adaptability aspect of the GAs. A very important
attainment is overcoming most of the problems arising in
techniques based on CCAT by allowing the GA.
Moreover, an enhancement in the performance of the
developed GA was achieved by applying the new
USPS crossover operators, which greatly improved the
convergence rate of the whole system.
REFERENCES
[1] A. Ahmadyfard and V. Abolghasemi, “Detecting license plate
using texture and color information,” in Proc. Int. Symp.
Telecommun., 2008, pp. 804–808.
[2] G. Li, R. Yuan, Z. Yang, and X. Huang, “A yellow license plate
location method based on RGB model of color image and texture
of plate,” in Proc. 2nd Workshop Digit. Media Its Appl. Museum
Heritages, 2007, pp. 42–46.
[3] X. Shi, W. Zhao, Y. Shen, and O. Gervasi, “Automatic license
plate recognition system based on color image processing,” in
Lecture Notes on Computer Science, Berlin, Germany: Springer-
Verlag, 2005, vol. 3483, pp. 1159–1168.
[4] M. Deriche, “GCC license plates detection and recognition using
morphological filtering and neural networks,” Int J. Comp. Sci.
Info Security, vol. 8, no. 8, pp. 263–269, Dec. 2010.
[5] O. Villegas, D. Balderrama, H. Dom´inguez, and V. Sa´nchez,
“License plate recognition using a novel fuzzy multilayer neural
network,” Int. J. Comput., vol. 3, no. 1, pp. 31–40, 2009.
[6] S. H. M. Kasaei, S. M. M. Kasaei, and S. A. Monadjemi, “A novel
morphological method for detection and recognition of vehicle
license plate,” Amer. J. Appl. Sci., vol. 6, no. 12, pp. 2066–2070,
2009.
[7] A. Theja, S. Jain, A. Aggarwal, and V. Kandanvli, “License plate
extraction using adaptive threshold and line grouping,” in Proc.
ICSPS, Jul. 2010, vol. 1, pp. 211–214.
[8] P. Tarabek, “Fast license plate detection based on edge density
and integral edge image,” in Proc. Int. Conf. Appl. Mach.
Intell. Inform.,2012, pp. 37–40.
[9] V. Abolghasemi and A. Ahmadyfard, “A fast algorithm for license
plate detection,” in Proc. Int. Conf. Visual Inform. Syst., 2007,
vol. 4781, pp. 468–477.
[10] S. Roomi, M. Anitha, and R. Bhargavi, “Accurate license plate
localization,” in Proc. Int. Conf. Comput. Commun. Electr.
Technol., 2011, pp. 92–99.
[11] S. K. Kim, D. W. Kim, and H. J. Kim, “A recognition of vehicle
license plate using a genetic algorithm based segmentation,” in
Proc. Int. Conf. Image Process., 1996, vol. 1, pp. 661–664.
[12] J. Xiong, S. Du, D. Gao, and Q. Shen, “Locating car license plate
under various illumination conditions using genetic algorithm,” in
Proc. ICSP,2004, vol. 3, pp. 2502–2505.
[13] Z. Ji-yin, Z. Rui-rui, L. Min, and L. Yinin, “License plate
recognition based on genetic algorithm,” in Proc. Int. Conf.
Comput. Sci. Software Eng., Dec. 2008, vol. 1, pp. 965–968.
[14] V. P. de Arau´jo, R. D. Maia, M. F. S. V. D’Angelo, and
G. N.R. D’Angelo, “Automatic plate detection using genetic
algorithm,” in Proc. 6th WSEAS Int. Conf. Signal Speech Image
Process., Sep. 2006, pp. 43–48.
ISBN NO : 378 - 26 - 13840 - 9
www.iaetsd.in
57

Iaetsd an efficient way of detecting a numbers in car

More Related Content

What's hot (20)

Viewers also liked (20)

Similar to Iaetsd an efficient way of detecting a numbers in car (20)

More from Iaetsd Iaetsd (20)

Recently uploaded (20)

Iaetsd an efficient way of detecting a numbers in car