Professional Documents
Culture Documents
T
Abstract: The license plate detection has become an important available which are different in terms of contrast and
application of Intelligent Transportation Systems. This can be background in. The license plates can be deliberately
used for a broader range of applications like toll collection, altered in fraud situations or replaced (e.g., with a stolen
reduce congestion and pollution. Since license plates can be plate), which is not possible with the VIN. The technique of
ES
replaced, stolen or simply tampered, they are not the ultimate
answer for vehicle identification. The objective is to develop a
system where by Vehicle Identification Number (VIN) is
digitally photographed and then identified by segmenting the
characters from the images captured. In this paper a novel
approach for segmentation of characters of VIN using
watershed algorithm is presented. Initially the VIN was kept in
identifying VIN is new and more challenging than license
plate recognition. This is because license plates have a fixed
format (at least) region-wise, while VIN number differs
according to the make and the model of the vehicle. Further,
the license plates have a standard and a visible look with
foreground and background with stark contrast, making
character segmentation comparatively easy. However, the
focus and images were captured. The images were then VIN embossed on the car body and that placed at the
subjected to pre-processing which consists of image processing dashboard have similar foreground and background
algorithms. These images were further efficiently processed by
contrasts which makes it difficult to decipher.
considering the distance transforms. The elapsed time and
II. RELATED WORK
A
entropy results give an analysis for increasing the efficiency
and good performance of watershed segmentation. In the past decades the telematics industry has increased in
size and scope so that technology developed in this field is
used to perform tasks ranging from updating accounts on
Keywords: Vehicle Identification Number; Distance toll roads to tracking vehicle progress for onboard
Transforms; Watershed Algorithm navigational systems. The scope of vehicle telematics seems
to be increasing where in VIN provides information on
IJ
where and how people are driving, what toll roads they are
I. INTRODUCTION using and what features are adding to their cars. This
information could be used as a powerful safety resource.
The Vehicle Identification Number (VIN) is a unique
The reason that VIN is popular for telematics purposes is
identification number for every car manufactured. A
that each car’s VIN is unique and that it can be decoded to
registration number can be changed, but the VIN is constant
give basic information concerning the car. Furthermore it is
throughout the lifetime of the car. Moreover, the registration
a permanent identifier and cannot legally be altered by sale
number is not unique to a car since it can be carried forward
or repair. Most cars also have their VIN in bar code form
to another car. The Vehicle Identification Number can be
which means it has the potential to be used by remote sensor
found most likely close to the engine, on the body frame,
systems as a tracking identifier. A number of techniques to
depending on the make or brand of the car. The wall behind
segment each character after localizing the license plate in
the motor (firewall) separates the engine bay and the
the image have been developed, such as feature vector
vehicles dashboard in the cabin. The VIN, in almost all
extraction and mathematical morphology [6], and Markov
cases, is in the middle, at the top of the firewall and either
random fields (MRFs) [7]. The work in [6] indicates that the
stamped in the same colour as the paintwork or on silver
method could be used for character segmentation in plates
“credit card” sized plate. The other types of VIN are also
with indistinguishable characters during off-line operation,
but since the algorithm is computationally complex it cannot known that which pixel belongs to which object. The image
be proposed for real-time license plate recognition. The is parted into regions and we know the discontinuities as the
method in [5] was developed for license plate segmentation boundaries between the regions. The different types of
in video sequences. However, the segmentation results were segmentations are:
far from suitable for automatic character recognition. By
A. Pixel-Based Segmentation
projection method (PM), characters are segmented
according to their height and width ranges after their four Point-based or pixel-based segmentation is conceptually the
boundaries are determined. In essence, template matching simplest approach used for segmentation.
method (TMM) is another form of PM, but with a more
B. Edge-Based Segmentation
considered segmentation procession and a more precise
boundary division. As for clustering method (CM), the Even with perfect illumination, pixel based segmentation
license plate characters are segmented by a clustering results in a bias of the size of segmented objects when the
analysis algorithm in pattern recognition, which overcomes objects show variations in their gray values. Darker objects
disconnection of characters, but it has a big computational will become too small, brighter objects too large. The size
load and slow processing. variations result from the fact that the gray values at the
edge of an object change only gradually from the
Watershed transform, which can separate an image into background to the object value. No bias in the size occurs if
T
many homogeneous non overlapped closed regions, has we take the mean of the object and the background gray
been widely applied in image segmentation algorithms. values as the threshold. However, this approach is only
Many watershed algorithms have been proposed [7]–[11]. possible if all objects show the same gray value or if we
Vincent and Soille proposed a watershed algorithm using apply different thresholds for each object. An edge based
immersion simulations [7]. With sorting before the flooding segmentation approach can be used to avoid a bias in the
process and with priority queue, this algorithm is size of the segmented object without using a complex
ES
dramatically faster than any former ones. Beucher and thresholding scheme. Edge-based segmentation is based on
Meyer’s algorithm also uses immersion simulations [8], [9]. the fact that the position of an edge is given by an extreme
Two types of algorithms are included: one creates watershed of the first-order derivative or a zero crossing in the second-
pixels and the other produces a complete tessellation of an order derivative.
image. An ordered queue is used in this algorithm, whose C. Region-based Segmentation
concept is similar to that of Vincent and Soille’s algorithm;
These methods focus attention on an important aspect of the
however, the minima of the input image need to be detected
segmentation process missed with point-based techniques.
and labeled first, thus increases the complexity of this
There a pixel is classified as an object pixel judging solely
algorithm. Dobrin et al.. proposed a fast watershed
on its gray value independently of the context. This meant
algorithm named split-and-merge algorithm [10]. It can
that isolated points or small areas could be classified as
solve the isolated area problems of the former two
algorithms when they are employed to create watershed object pixels, disregarding the fact that an important
A
pixels. Although the results obtained are more correct, it is characteristic of an object is its connectivity. If we use not
the original image but a feature image for the segmentation
more complex than the other two algorithms. Moreover,
process, the features represent not a single pixel but a small
watershed transform for video segmentation is often
neighbourhood, depending on the mask sizes of the
required to produce a tessellation of an image, where the
operators used. At the edges of the objects, however, where
isolated area problem would not occur. Moga et al..
the mask includes pixels from the object and the
proposed a watershed algorithm suitable for parallel
background, any feature that could be useful cannot be
IJ
T
( ) *| |,| |} (4)
management of data; another example of multimedia
applications is the dissemination of information in the
network .Today, a large number of multimedia data streams B. Watershed Algorithm
sent on the Internet, However, due to the bandwidth The idea of watershed is drawn from a topographic analogy.
limitations; we need to compress the data, and therefore it Quite naturally the first algorithm for computing watersheds
calls for image and video segmentation. is found in the field of topography [13]. The introduction of
ES
1.) Region Growing Algorithm and Watershed Algorithm
The basic idea of region growing method is a collection of
pixels with similar properties to form a region. The steps are
as follows:
(a) Find a seed pixel as a starting point for each of needed
the watershed transformation as a morphological tool is due
to Digabel et al [1]. Watershed is then approached
theoretically by F. Maisonneuve and used in numerous
grayscale segmentation problems. Currently, it is being
studied from theoretical, practical, and algorithmic points of
view. The watershed transform applied to the image does
not produce contours of the features. On the contrary, it
segmentation.
partitions the image into the associated areas by the intensity
(b) Merge the same or similar property of pixel (Based on a gradient and considers the gradient image as a topographic
pre-determined growing or similar formula to determine) relief, where the intensity of a pixel denotes the altitude of
with the seed pixel around the seed pixel domain into the that pixel. Each pixel in this digital image is assigned a label
domain of seed pixel.
A
during the transformation of the catchments basin of a
(c) These new pixels act as a new seed pixel to continue the regional minimum. When finished, the resulting network of
above process until no more pixels that satisfy the condition dams defines the watershed of the image. Compared to the
can be included. other methods, the watershed has several advantages as
The watershed algorithm is more representative in the follows [2], [3].
application of mathematical morphology theory for image − The gaps are handling properly and the placement of
segmentation. Watershed algorithm is a region based boundaries is at the most significant edges.
IJ
segmentation techniques image that uses image morphology. − The resulting boundaries form closed and connected
Watershed algorithm is an iterative adaptive threshold regions.
algorithm. The idea of watershed algorithm is from Watershed algorithm (WA) is a widely-used method in
geography, it see gradient magnitude image as a topographic image segmentation and image edge detection. Generally, it
map, the gradient magnitude in correspond with altitude, the is used to process the gray gradient image. Its basic thought
different gradient in correspond with the peak and basin in is that the image is regarded as topology geomorphology in
valley in the image. It sees every object of image (including Geodesy and each pixel value in the image is taken as the
background) as a separate part and requested there must altitude above sea level. Each local minimum in the image,
have one tag at least in the each object (or seed points). including its neighbourhood (namely, Influence Zone in
Geodesy), is called catchment basin, and the boundary of the
IV. BASIC DEFINITION catchment basin is called watershed. The principle of WA is
shown in Fig. 1. In fact, the Essence of WA is that finds out
A. Basic definition the local maximum in the segmented region.
Definition1. Suppose S is a connected region and B is the
boundary of the region S. The region area A is defined as
T
the set of the coordinate ( ) and f ( )<T, namely: are used to connect the objects together to indicate the flow
of data between them. The data that we enter into the front
)= { ( )| f ( )<T} (5) panel controls enter the block diagram through these
( )(
terminals and after execution the output data flow to
( ) is the set of the pixels in catchment basins when the indicator terminals where they exit the block diagram, re-
gray value is T , then: enter the front panel and appear in front panel indicators
( )=⋃ *S( )⋂ (
(7)
giving us final results.
Test image1
T
The noise has been removed using filtering techniques of
above 3 parts individually .the filter has been applied on
each pixel location of fig.2. The filter is more powerful than
remaining filters like spatial filters. With the filter, we
choose one of two filtering modes: correlation or
ES convolution can be performed. The difference between the
two is that convolution rotates the filter by 180 o before
performing multiplication. The filter used has been obtained
from
( )
⁄
( ) ( ) (5)
( )
( )
∑ ∑
(6)
Figure2. Isolated regions using LabVIEW
VII. BINARIZATION
A
After pre-processing of image it is converted in to binarized
VI. PRE-PROCESSING form. Here the binarization is such that the VIN numbers are
white and the pixel is 1; background is black and the pixel is
The other images that have been considered for processing 0.If any noise occurs in background image then the
are critical and this can be used on various images. The pre- morphological operation is done to eliminate the noise. The
processing of these images is difficult and quite complex. In foundation of morphological processing is in the
pre-processing the image, first split the VIN number image mathematically rigorous field of set theory; however, this
IJ
(Fig.3) in to 3 parts. Because it includes darker background level of sophistication is seldom needed. Most must be the
in last 6 VIN numbers and lighter background in first 6 VIN end of a line, and therefore shouldn't be erosion of the thick
numbers. The standard VIN number image in fig. ridges takes place from the outside. In other words, if a pixel
3considered is 138*865.The second image is an example of is black, and it is completely surrounded by black pixels, it
images which are quite common but it also requires pre- is to be left alone on this iteration. Running the algorithm
processing as it has noise. both ways shows that it works better. Remember, this is
very common in morphological image processing; trial and
error is used to find if one technique performs better than
another. The pixel must have more than one black neighbor.
If it has only one, it is changed in to a continuous line, not a
group of interrupted segments. As show in fig.4 is the
binarized image.
T
Figure4. Binarized Image Figure 5.Euclidean Distance
T
Chessboard Distance Transform might be unable to separate
Figure 8.Watershed and Inverse Watershed for Chess Board Distance
the overlapped components.
REFERENCES
[1] H. Digabel, C. Lantuejoul, “Iterative algorithms,” In: Proceedings of
Quantitative Analysis of Microstructures in Material Science, Biology and
Medicine. West Germany: Riederer Verlag, pp. 85–99, 1978.
[2] P.R. Hill, C.N. Canagarajah, D.R. Bull, “Image Segmentation Using a
Texture Gradient-Based Watershed Transform,” IEEE Trans. Image
Processing, pp. 1618–1633, 2003, 12.
[3] H.T. Nguyen, M. Worring, R.V.D. Boomgaard, “Watersnakes: Energy-
Driven Watershed Segmentation,” IEEE Trans. Pattern Analysis and
T
Machine Intelligence, pp. 330–342, 2003, 25.
[4] Y. Cui and Q. Huang, ”Extracting characters of license plates from
video sequences,” Mach. Vis. Appl., vol. 10, no. 5/6, pp. 308-320, Apr.
1998.
[5]Z. G. Han, S. Y. Lao, Y. X. Xie, et al, “License plates segmentation and
adjustment,” Computer Engineering and Applications, vol. 9, pp.210-212,
2003.
[6] S. Nomura, K. Yamanaka, O. Katai, H. Kawakami, and T. Shiose, ”A
ES novel adaptive morphological approach for degraded character image
segmentation,” Pattern Recognit., vol. 38, no. 11, pp. 1961-1975, Nov.
2005.
[7] Y. Cui and Q. Huang, ”Extracting characters of license plates from
video sequences,” Mach. Vis. Appl., vol. 10, no. 5/6, pp. 308-320, Apr.
1998.
[8] F. Meyer, “Color image segmentation,” in Proc. Int. Conf. Image
Processing and Its Applications, 1992, pp. 303–306.
[9] E. R. Dougherty, Ed., Mathematical Morphology in Image Processing.
New York: Marcel Dekker, 1993, ch. 12, pp. 433–481.
[10] B. P. Dobrin, T. Viero, and M. Gabbouj, “Fast watershed algorithms:
analysis and extensions,” in SPIE Nonlinear Image Processing V, vol.
2180, 1994, pp. 209–220.
[11] A. Moga, B. Cramariuc, and M. Gabbouj, “An efficient watershed
A
segmentation algorithm suitable for parallel implementation,” in Proc. Int.
Figure 10.Watershed and Inverse Watershed for City Block Distance Conf. Image Processing, 1995, pp. 101–104.
[12]J. B. T. M. Roerdink and A. Meijster, “The watershed transform:
A clear distinction can be observed between the watershed Definitions, algorithmsand parallizations strtagies,” Fundamental
informaticae, vol. 41, pp. 187-228, 2000.
segmentation being applied to different distance transforms [13]S.H. Collins, “Terrain Parameters Directly from a Digital Terrain
from figs 8, 9, 10. Model,” Canadian Surveyor, pp. 507-518, 1975, 29.
IJ