NUMBER PLATE DETECTION SYSTEM

Anant Bishnoi1, Kavita Saini2

1.
B.Tech, Department Of Computing Science Of Engineering,
Galgotias University, Greater Noida, India
bishnoianant9@gmail.com
2
. Associate Professor,
Galgotias University,India

Abstract: The aim of this research paper is to style and analyze the number Plate recognition systems program
mediated through Digital Images, especially by using desktop peripherals. In doing so, license plates attached to
cars are becoming to be the test subject of this research paper. the basic goal of the ANPR program itself is
essentially to utilize a digital image identification system so on spot every single vehicle that goes in and out
through the barrier gate of the parking system. However, the social scenario in India is significantly different thanks
to problems like poverty, unemployment also as a considerably lower respect for rules. This makes it infeasible to
travel for a totally automatic tollbooth. The industry requires an automatic vehicle arrangement in India to not
reduce or eliminate human intervention or labour, but to create sure that human intervention doesn't cause any
financial malpractices. The industry requires a system that runs within the background and merely keeps a cross-
check on the manual.This has been written in python artificial language . Many applications are often implemented
by using this system like security, highway speed detection, violation of sunshine , identification of handwritten text,
discovery of stolen cars, automatic fee collection system.

I.Introduction techniques used for vehicle identification purposes.

Toll Booths in India generally employ a purely
sensory system of auto classification. However, this
causes a large loss of revenue to the firms operating
the toll booths thanks to rampant malpractices and
discrepancies. to stay a tab on the operators some
tollbooths employ a system using fibre optic sensors
to automatically classify a vehicle within the
background and tally the results with the manual
entries. However, this method is dearly complicated
and requires high maintenance. We aim to review the
many systems which might be accustomed replace
such a system with a less expensive and efficient
alternative.
As already stated, the system using fibre optics
inherently possesses many problems apart from the
foremost concerns of high cost and maintenance.
Although an IR curtain system reduces the value
significantly, it's still quite expensive and cheaper
alternatives are desired.As most toll booths employ
cameras for security purposes, it's been felt that the
feasibility of a system using IP cameras should be
tested.
the sole intention of this project is to hunt out the
foremost efficient thanks to recognizing the
registration information from the digital image
(obtained from the camera). This process usually
comprises three steps. initiative is that the registration
number plate localization, no matter the license-plate
size and orientation. The second step is that the
segmentation of the characters and also the last step
is that the recognition of the characters from the
registration code . Thus, this project uncovers the
essential idea of varied algorithms that are required to
accomplish character recognition from the
registration number plate during Template Matching.
This above mentioned feature of the algorithm helped
in achieving faster character recognition of the
registration code . The steps involved during this
process of character recognition are: Image
processing, Defragmentation, Resizing and Character
localization that are required to be performed on the
image for Template Matching to be done.

Number Plate recognition is one amongst the

 Loading a picture

2. Grayscale conversion:
II. Data flow diagram for ANPR system:
cvCvtColor() function is used to convert this
RGB image to grayscale image.

Fig.4
Grayscale Image

Steps followed:
3. Histogram equalization:
Number plate recognition basically consists of three Histogram equalisation is performed on this
concrete steps mentioned below: image using cvEqualizeHist() function.
1. Number Plate Extraction.
2. Character Segmentation.
3. Template Matching.

1. Loading an RGB image:

The image whose number plate recognition

is to be done is loaded.

Fig.5
Histogram Equalized Image

Binarization:

This image is then converted to binary by the use

Fig.3 of adaptive thresholding.
 Fig.6
Fig.8
Binarized Image
Edge Detected Image

6. Plate region extraction:

4. Dilation:
Plate region is acknowledged by passing an
This image is dilated using the cvDilate() oblong image over the previous using
function. cvMatchTemplate() function

Fig.9
Extracted Plate Region

7. Character segmentation:

Characters are segmented from the number

plate image which is then used for template
matching.

Fig.7
Dilated Image

Fig.10
Segmented Character

5. Edge detection: 8. Template matching:

Dilated image is subtracted from the

original image to get the edges.
 Segmented characters are template matched
with the templates of each character and the
number plate is identified as a string. IV.REFERENCES

Papers:

1. S. HamidrezaKasaei, S.
MohammadrezaKasaei, S.
AlirezaKasaei
2. International Journal of Computer
Theory and Engineering, Vol. No.
2,
3. 2 April, 2010.

Fig.11 4. Serkan Ozbay, and Ergun Ercelebi

Output in string form 1793-8201
5. World Academy of Science,
Engineering and Technology 9
2005

6. Deepak Kumar Gupta-Y6154,

Siddhartha Kandoi-Y6472
Fig.12 7. CS 676: Image Processing and
Output after Template Matching Computer Vision 2009-10
Semester 1
III.CONCLUSION
This particular writing elaborates a series of
experimentation that we have conducted with Websites:
different conditions and problems that occurred
within the process of digital projection on the license 1. http://www.anpr-tutorial.com
plate as its object. Our system results show 99.99%
accuracy for high quality image 80.6% for medium 2. http://www.cctvinformation.co.uk/i/An_Intr
quality and 47.5% for low quality image, this result oduction_to_ANPR
is still lower than previous work and needs to be
improved. 3. http://www.visl.technion.ac.il/projects/2003
w24/

4. http://www.stackoverflow.com

5. http://opencv.willowgarage.com/documentat
ion