Enhanced Algorithm For Tracking number Plate From Vehicle Using

IJournals: International Journal of Software & Hardware Research in Engineering
ISSN-2347-4890
Volume 3 Issue 4 April, 2015
Enhanced Algorithm For Tracking number Plate
From Vehicle Using Blob Analysis
Author: Shashikumar Naubadkar1 ; Dr. A Meera2
PG Scholar1 ECE Department, BMSCE Bengaluru ; Professor2 ECE Department, BMSCE Bengaluru
shashiln5051@gmail.com1 ; amira.ece@bmsce.ac.in2
ABSTRACT
This paper presents an enhanced algorithm to extract the
text from the number plate of the vehicles. The enhanced
algorithm is based on Prewitt Edge Detector and text
extraction is carried out by BLOB analysis. Experiments
are done on many number plates and results show that it
has overcome the problems for many problematic images,
where text was not extracted using old methods because
either the system does not extract number plate from gray
scale due to some luminance conditions or due to
problematic backgrounds. Extracted text is stored in text
file.
Keywords–number plate
detector, BLOB analysis
extraction,
prewitt
edge
1. INTRODUCTION
Number plate extraction is hotspot research area in the
field of image processing. Many of automated system
have been developed but each has its advantages and
disadvantages. Previously it was was assumed that the
old method worked on images which have been captured
from fixed angle parallel to horizon in different luminance
conditions. It was also assumed the vehicle is stationary
and images are captured at fixed distance.[1] developed
algorithm which is applied on the car park systems to
monitor and manage parking services. Algorithm is
developed on the basis of morphological operations and
used for number plate recognition. Optical character is
used for the recognition of characters in number plate. [2]
proposed a methodology which finds ROI using
morphological processing and directional segmentation.
The ROI is the area which includes the number plate from
which alphanumeric characters are recognized.
2.PROPOSED ENHANCED ALGORITHM
An automated system is developed using MATLAB in
which image is captured from camera and converted in
Gray scale image and then to binary image for pre
processing. After conversion, Prewitt edge detector is
© 2015, IJournals All Rights Reserved
applied then dilation process is applied on image and
unwanted holes in image have been filled. After dilation,
blob analysis is applied which filter out unwanted regions
or unwanted noise from image, and image is segmented.
After segmentation, each alphanumeric character on
number plate is extracted and then recognized with the
help of template images of alphanumeric characters. Each
alphanumeric character is stored in file and whole number
plate is extracted successfully.
Figure 1. block diagram of the proposed algorithm
Enhanced algorithm that uses Prewitt edge detector [3 ]
and blob analysis
1. Convert the image into monochrome image by
thresholding.
2. Filter the image for removing noise. Use Gaussian lowpass filter.
3. Apply Prewitt edge-detector to the filtered image.
4. Apply proper morphological operations, i.e. dilation to
make clusters of text regions.
5. Apply blob analysis to segment and extract the text
from the numberplate
6. Apply segmented and extracted text image to OCR to
convert into .txt file.
3. BLOB ANALYSIS
The Blob Analysis block is used to calculate statistics for
labeled regions in a binary image. The block returns
quantities such as the centroid, bounding box, label
matrix, and blob count. The Blob Analysis block supports
input and output variable size signals.
www.ijournals.in
Page 85
IJournals: International Journal of Software & Hardware Research in Engineering
ISSN-2347-4890
Volume 3 Issue 4 April, 2015
information and ignore the rest.
Area of a BLOB :It is the number of pixels the BLOB
consists of. This feature is often used to remove BLOBs
that are too small or too big from the image.
Figure 2.blob analysis block
Count is the scalar value that represents the actual
number of labeled regions in each image.BW is the vector
or matrix that represents a binary image. Area is the
vector that represents the number of pixels in labeled
regions. BBox (Bounding box) is the M-by-4 matrix of [x
y width height] bounding box coordinates, where M
represents the number of blobs and [x y] represents the
upper left corner of the bounding box. Example: Suppose
there are two blobs, where x and y define the location of
the upper-left corner of the bounding box, and w, h define
the width and height of the bounding box. The block
outputs at the BBox port.
Bounding box of a BLOB :It is the minimum rectangle
which contains the BLOB, see Figure 1. It is defined by
going through all pixels for a BLOB and finding the four
pixels with the minimum x-value, maximum x-value,
minimum y-value and maximum y-value, respectively.
From these values the width of the bounding box is given
as
w = Xmax – Xmin
and the height h= Ymax – Ymin.
A bounding box can be used as a ROI (region of interest).
Figure 4. Bounding box of letter G
First we have to separate the different objects in the image
and then we have to evaluate which object is the one we
are looking for. The former process is known as BLOB
extraction and the latter as BLOB classification.
A BLOB consists of a group of connected pixels.
The term large indicates that only objects of a certain size
are of interest and that small binary objects are usually
noise. A number of different algorithms exist for finding
the BLOBs. One of these algorithms known as the Grassfire algorithm, we use 4-connectivity for simplicity. The
Recursive Grass-Fire Algorithm starts in the upper-left
corner of the binary image. It then scans the entire image
from left to right and from top to bottom.
Bounding box ratio of a BLOB :
It is defined as the height of the bounding box divided by
the width. This feature indicates the elongation of the
BLOB, i.e., is the BLOB long, high or neither.
Compactness of a BLOB: is defined as the ratio of
the BLOB‟s area to the area of the bounding box. This can
be used to distinguish compact BLOBs from noncompact
ones. For example, fist vs a hand with outstretched
fingers.
Compactness =
Area of BLOB/ Width * Height\
4. OPTICAL CHARACTER ECOGNITION
OCR, in principle classify optical pattern corresponding
to alphanumeric or other characters. OCR converts
scanned image of text - printed or hand-written into
machine-editable text format electronically. In simple
terms, it refers to the conversion of images of handwritten, typewritten or printed-text usually taken by
means of scanner or a suitable camera into machineeditable form [4].
Figure 3. 4-connectivity
BLOB Extraction : The purpose of BLOB extraction is to
isolate the BLOBs (objects) in a binary image. Feature
extraction is a matter of converting each BLOB into a few
representative numbers. That is, keep the relevant
© 2015, IJournals All Rights Reserved
Initially a library has been created with 26 uppercase
and 26 lowercase alphabets and 10 numeric bitmaps.
These bitmaps are binary replica of all the alphanumeric
characters and each of them is stored in matrix form. They
all are of equal size and dimension. Figure.4.2 shows the
bitmap stored in library of X.
www.ijournals.in
Page 86
IJournals: International Journal of Software & Hardware Research in Engineering
ISSN-2347-4890
Volume 3 Issue 4 April, 2015
Figure 5. Bitmap image corresponding to letter „X‟
5. SIMULATION RESULTS
The algorithm implementation was carried out on
MATLAB platform. Figure 6 shows the result of the
proposed algorithm.
(d)
(a)
(b)
(c)
(e)
Figure. 6. text extraction from the number plate
(a)Original image (b)histogram of original
image
(c)extracted text (d) extracted text is segmented
(e)extracted text in .txt file
Figure.6 shows the text extraction from the numbetplate.
(a) is the original color image of a car. (b) is the histogram
of the color image, the right hand side has more
frequency, represents light and pure white areas, the
image has. (c) is the extracted text from the original
image.(d) extracted text is segmented and (e) applied to
OCR and saved in text file.
Figure.7 shows comparision of the proposed algorithm
for complex number plate where text was not extracted
by [5].The proposed algorithm is successful in extracting
the text from the complex number plates.
(a)
© 2015, IJournals All Rights Reserved
www.ijournals.in
Page 87
IJournals: International Journal of Software & Hardware Research in Engineering
ISSN-2347-4890
Volume 3 Issue 4 April, 2015
[6] Mrs. Neha Gupta, and Mr.V .K. Banga , “Image
Segmentation for Text Extraction” April 2012, 2nd
International Conference on Electrical, Electronics and
Civil Engineering (ICEECE'2012) Singapore.
(b)
(c)
Figure.7 comparision of the proposed algorithm for
complex number plate
(a) original image (b) text not extracted by [5]
(c) text extracted by proposed algorithm
6. CONCLUSION
The paper has implemented an enhanced algorithm for
text extraction from complex number plate and tested on
with nearly 20 images. The results shows that the
proposed algorithm has overcome the problems with the
existing methods.
[7] Prof. Amit Choksi1, Nihar Desai2, Ajay Chauhan3,
Vishal Revdiwala4, Prof. Kaushal Patel, Electronics and
Telecommunication Department, BVM Engineering
College, Anand, India, “Text Extraction from Natural
Scene Images using Prewitt Edge Detection Method ”,
International Journal of Advanced Research in Computer
Science and Software Engineering Research Paper,
Volume 3, Issue 12, December 2013.
[8] C.P. Sumathi1, T. Santhanam2 and G.Gayathri Devi
“A SURVEY ON VARIOUS APPROACHES OF TEXT
EXTRACTION IN IMAGES” International Journal of
Computer Science & Engineering Survey (IJCSES) Vol.3,
No.4, August 2012 .
[9] Yungang Zhang and Changshui Zhang “A New
Algorithm for Character Segmentation of License Plate”.
[10] Hyung Il Koo, Member, IEEE, and Duck Hoon Kim,
Member, IEEE “Scene Text Detection via Connected
Component Clustering and Nontext Filtering” IEEE
TRANSACTIONS ON IMAGE PROCESSING, VOL.
22, NO. 6, JUNE 2013
7. REFERENCES
[1] J.S. Chittode and R. Kate, “Number plate recognition
using segmentation,” International Journal of Engineering
Research & Technology, Vol. 1 Issue 9, November- 2012.
[2] C N Paunwala and S Patnaik, ”A novel multiple
license plate extraction technique for complex background
in Indian traffic conditions,” International Journal of
Image processing, Vol-4,Issue-2,pp 106-118.
[3] Prof. Amit Choksi,Nihar Desai, Ajay Chauhan “Text
Extraction from Natural Scene Images using Prewitt
Edge Detection Method”.
[4] Disha Bhattacharjee, Deepti Tripathi “A Novel
Approach for Character Recognition” International
Journal of Engineering Trends and Technology (IJETT) –
Volume 10 Number 6-Apr 2014.
[5] Manisha Rathore and Saroj Kumari “TRACKING
NUMBER PLATE FROM VEHICLE USING
MATLAB” International Journal in Foundations of
Computer Science & Technology (IJFCST), Vol.4, No.3,
May 2014.
© 2015, IJournals All Rights Reserved
www.ijournals.in
Page 88