2015 International Conference on Circuit, Power and Computing Technologies [ICCPCT]

DEVELOPMENT OF REAL TIME VEHICLE ENTRY

MONITORING SYSTEM
Annet Ronika.D,
Dept.of ETCE, Sathyabama University, M.S.Godwin Premi
Chennai Dept.of ETCE, Sathyabama University,
[email protected] Chennai

Abstract: Vehicle monitoring or management system is used years. Automatic vehicle identification systems are used for an
to identify vehicles approaching ‘manned’ & ‘unmanned’ effective process. Automatic number plate recognition is one
Check-Posts at entry/exit with real-time monitoring at the of those methods that perform pre-processing over the images
central control room. This paper proposes an entry obtained from the imaging sensor prior to transmission. The
monitoring system which is mainly focused for university band width limitations rules out the possibility of the direct
campus. The projected system contains a video camera transmission of videos from the imaging sensors or cameras.
which is placed in front of the check post and is controlled To indentify the vehicles license plate image process
by a sensor. The video camera which is used can capture technology is used. Entire algorithm used for vehicle number
slow motion video whenever the check post is raised. The plate prior to the application of OCR technique with different
video obtained from the camera is processed and the vehicles basis of attributes associated with license plate. The system
are recognized from their number plates. Image processing adopted the optical recognition to read the license plate from
technique is used to recognize the number from the license the images based on extraction of license plate region,
plate. The vehicles are identified from the database whether recognition of character and segmentation of plate characters.
the vehicles belongs to university or unknown. The complete Template matching and training which has been frequently
data including vehicle information, timestamp and Check used methods. The previous work has included the both
Post details are transmitted in real-time to the central control approaches for vehicle detection and character recognition and
room that allows authorities to monitor the vehicles. At the also include character segmentation but in the proposed work
central receiver full vehicle entry records are stored and if was involved with recognition by perception. This paper aims
any unknown vehicle entry is detected it is easily monitored determination of the license plate is recognition by perception
by the central controller. efficiently over real time.

II. EXISTING SYSTEM

Keywords: slow motion video, real-time monitoring,
verification vehicle number plate Images are captured and it was compensated for noise and
illumination variations. In the existing system the vehicle’s
number plate have been designed for character recognition and
I. INTRODUCTION character segmentation. The vehicle identification is employed
to recognition through the genetic algorithm. Features based
Nowadays vehicles play a vital role in transportation. Vehicle matching is adopted for recognition. The vehicle’s
identification is an essential stage in an intelligent traffic identification by recognition technique consists of plate loc
system. Object detection is still a challenging task it was a alization, character recognition and vehicle detection. The
critical part in many applications like image understanding, method recognition linear to the number of object is presented
scene analysis and image search. Modern transport system is in the existing work. Edge and gradient techniques are used to
highly intelligent and automatic. Automatic identification of a extract for the ground details. They consists of plate character
vehicle has become a critical component of the intelligent extraction, character recognition and vehicle detection. The
transport license plate recognition. The general algorithm gray level transitions images were computed in the object
automatic number plate recognition system was presented to regions. A variation of algorithm to detect for shape and edge
detect the vehicles violating traffic signals. The use of vehicle .Based on the geometrical features of the detected window
has been increasing because of the population growth in recent such as height width and aspect ratio objects are categorized

978-1-4799-7075-9/15/$31.00 ©2015 IEEE

 2015 International Conference on Circuit, Power and Computing Technologies [ICCPCT]

as small medium and heavy vehicles. They analysis through algorithm to indicators to performance detection accuracy and
character recognition and character segmentation. The vehicle detection time. Standard features vector for all characters are
number plate which can be fixed at the bottom or both in front computed and maintained in a database. Vehicle identification
and back. Either the characters are printed in single row in two by recognition is designed and tested for vehicles number
rows. Each character is uniquely for recognition. Actual plate have been tested in different light conditions, diverse
recognition of extracted character is combining the various angles in the system .It could detect the vehicles characters in
features extracted in pervious stages to give actual text in the the number plate quickly recognize with good accuracy. The
images. Characters are normalized is to filter the characters proposed system can be applied with required modification by
into a block containing no extra white spaces. Template recognition for perception.
matching has been a successful algorithm for recognition of
characters. The character image is compared with database Table1. Existing work table
which is already stored in exists in the database .The output of
the segmentation stage is considered the characters contained EXPERIMENTAL RESULT OF EXISTING WORK
in the image were compared character appearing in the Test Success Failure Detection/
Stages
number plate image. Character segmentation deals with the image instance instance recognition
separation of text pixels from the background pixels. The Detection 108 100 8 92.5
output of the step in a binary image where black text character Recognition 310 283 27 91
appear on a white background. License plate characters are
segmented into parts obtaining the characters individually.
Image is filtered for enhancing the image and removing the III.PROPOSED SYSTEM
noises and unwanted spots .The extraction of actual text
Moving vehicles generate different patterns depending on their
regions diving pixels with similar properties into contour or
working conditions and information necessary to classify the
segments portions of the frames.
vehicles. The processing time is proportional to the size of the
input images which will reduce the image size by performing
image normalization. The image has normalized resolution. As
the working process of the proposed system have been started
sensor signal sense signal from the camera which will
capture video of the vehicle number plate .the license plate
video processing is converted into frames . As the frames can
be obtained from a video and converted into images .median
filter used for removal of noise these filter out the unwanted
noise in back ground whose number plate. Analysis number
plate detection by character recognition and character
segmentation to recognize the license plate. The image is
loaded to the morphological processing involves rgb to gray
scale conversion which is include with gray scale into binary
images are extracted from morphological process Images are
Fig 1.Example of character segmentation
resizing through the pixel values by converting through the
frame conversion of the images .The vehicles number plate
Pattern recognition which may be a higher interest to
has been stored and also compared with database which
formalize and visualize the pattern is the assignment to the
already exists in the database. While entering the number plate
input value. As recognition system must contain some
it will detection recognition by perception once the vehicle
memory of the object to recognize to match with stored image.
number plate stored in the database it need not over write in
The characters are matching due the template matching which
database if not saved in the database need to enter the vehicle
already stored the values of each character. The algorithm
number plate in the database it will stored as the output of
produces good result in detection accuracy and time. It was
recognition by perception. The proposed system has different
carried out with LPR images for the better result for edge
steps to process recognition by perception as discussed below
images obtained by gradient methods than threshold methods
it is employed to detect vehicle information. Any recognition
 2015 International Conference on Circuit, Power and Computing Technologies [ICCPCT]

Extraction of Plate Region Correlation

The extraction of number plate region could detect the Correlation in the image processing is an effective technique
character quickly with good accuracy. When the image has for image recognition has been implement in many vehicle
been analysis to captured from the camera is first converted recognition concepts. There are two types correlation auto and
into binary image consisting of only 1’s and 0’s by cross. Auto correlation purpose involves only one signal and
thresholding the pixel value of 0 for all pixel in the input provides information about the structure of the signal. Cross
image with luminance less than threshold value and 1 other correlation role shared properties between two signals.
pixels. Character are printed either in a single row or in two Unknown input image and known database image are
has study the features number plate is extracted. Captured occupies in this system.
image are converted into binary image. Using connected Localization
component analysis which could be density of line features As the localization stages included localizing the text in
Number plate is localized and also extracted in effective images for detection which could able to serves as the input
manner. Template matching and training which has been image and source image for the plate localization. The text
frequently used method. The video are captured from camera detection could phase seeks to detect the presence of text in a
those video containing text the rate could be increased or given image .Selected a frame containing text from elected by
decreased depending on the speed of those video based on the video framing is very low threshold value because the portion
frames per sec. Smearing method for extraction of text areas occupied by a text region .For the number plate detection or
on a mixed image. The horizontal and vertical smearing is localization techniques based on edge statistic and
applied for segmentation as to produces good accuracy result. mathematical morphology gives very good result.

Filtering Table 2. Proposed work table

Median filter is a non-linear process useful in reducing
EXPERIMENTAL RESULT OF PROPOSED WORK
impulsive or salt pepper noise .The filter technique to
Test Success Failure Detection/
performance removal noise from image under consideration Stages
image instance instance recognition
and also helps to removing the impulsive noise conserve the Detection 100 95 5 95
edges. As they consequence those extremes values with noise
Recogniti
which could lie as far as away from the median value are 300 287 25 90
on
removed by the filter. The most useful in this technique is
preserving edges in an image while reducing random noise. Recognition by Perception
Filter is more effective than convolution when the goals has This an important phase in the vehicle number plate when
simultaneously reduced noise and preserve edges. As the low image is captured in the camera. An images has been
pass filtering median filtering smoothes the image and converted into frames .The rbg images changes into gray
reducing the noise of the filter can preserve discontinuities A image dose not reduce the size of image. When the vehicle
few pixels values differ in a function for smooth images enters automatically the characters in number plate has been
significantly from their surrounding without affecting any register in database by recognition The pattern recognition of
other pixels the vehicle could recognition once without analysis signal
character as in the exiting method it only percept number
Morphological Processing plate once and stored automatically in database.
The morphological processing is an important role in any
image processing technique .It revolves around the conversion
of images and the filtering of the image processing. Median
filter is a non-linear filter which is useful in reducing
impulsive noise or salt and pepper noise in the images and it
also include removal of noise in edges.
The rgb images which has converted to gray scale images for
an improvement in the images those images whose converted
to binary image to have threshold value to define the images
have been extraction of the vehicle number plate.
Fig 2. Example of character by perception
 2015 International Conference on Circuit, Power and Computing Technologies [ICCPCT]

Storing and Comparing

This final stage of process in recognition for perception while
the vehicle entry in the check post the camera capture as video
by using sensor those video are converted into frames which
recognizes the character in the number plate those character
are saved in database .If the vehicle already saved in the
database need not save because it will compare with in the
stored data .If any new vehicle enters in the check post have
gives details of number plate to the database and save as
output result.

System Setup

Fig 4. Flow chart

Fig 3. System setup

Algorithm
The procedure of recognition for perception which takes place
in proposed work is given below.

Fig 5. Sample images captured from the camera

 2015 International Conference on Circuit, Power and Computing Technologies [ICCPCT]

Experiment Result for Proposed Work

Fig 8. Output result stored in the database

nplate= nplate=
TN0BT5213 TN12E4316

nplate= nplate=
AED632 OHZ2692
Fig 6. Color images of the number plate

Fig 9. Output image saved in the command window

IV. CONCLUSION
As far as the recognition by perception technique is much
more effectively used to identify the vehicles number plate
and stores in the database. This project has advantages are
highly perception, no wastage of resources and it can be
implemented at several time by the moving vehicle. The
sensor senses the signal to web camera once vehicle entry in
the check post which will capture the images and has been
converted into frames .The images is captured by the web
camera it is connected by usb cable. In the existing work
signal character are analysis which will take more time
consumption it’s not suitable in the high speed traffic. To
overcome these drawbacks the recognition by perception is an
effective method to examination the vehicles license plate.

REFERENCES
Fig 7. Binary image of the number plate
[1] P.Vijayalakshmi,M.Sumathi “Design of algorithm for
vehicle identification by number plate recognition” IEEE
fourth international conference ICOAC December 2012
MIT,Anna University Chennai Dec 13-15 2012
[2] Muhammad Junaid Muzammil,Sajjad Zaidi “Application
of image processing techniques for the extraction of
vehicle number plates over ARM target board”.
[3] Mrugesh.L.Shah,Prarthan.D.Mehta,“Classification of
vehicles using adaptive neuro fuzzy inference system
“2014.IEEEstudent’s conference on Electrical,Electronics
and Computer Science.
 2015 International Conference on Circuit, Power and Computing Technologies [ICCPCT]

[4] Huazhen Li,Changle Zhou,Wei Xue,Yinbin Guo “License

Plate Recognition Based On Intrinsic Image
Decomposition Algorithm”.The 9th International
Conference on Computer Science& Education( ICCSE
2014) August 22-24 2014.
[5] Jayshree Ghorpade,Raviraj Palvankar,Ajinkya
Patankar,Snehal Rathi,“Extracting Text from
Video”,Signal & Image Processing:An International
Journal (SIPIJ) ,vol2,no.2 June 2011.
[6] Jun Kong, Y.L., Xingu Liu, X. Zhou, 2005. “A novel
license plate localization method based on Textural
feature analysis”. Proceedings on IEEE International
Symposium on Signal Processing and Information
Technology, 275-279.
[7] Duan, T.D., T.L.H. Du, N.V. Hoang,” Building an
automatic vehicle license-plate recognition system”.
Proceedings of International Conference on Computer
Science in 2005,pp 59-63.
[8] R.Radha,C.P.Sumathi,“A novel approach to extract text
from license plate of vehicles”,An International Journal
(SIPIJ) vol 3,no.4,August 2012.
[9] Serkan Ozbay, Ergun Ercelebi “Automatic vehicle
identification by plate recognition”, World Academy of
Science,Enginnering & Technology ,International Journal
of ECE ,Robotics &EEE ,vol:1 no:9,2007.
[10] E.R. Lee, P.K. Kim, and H.J. Kim, 1994, “Automatic
Recognition of a Car License Plate Using Color Image
Processing,” Proceedings of the International
Conference on Image Processing.
[11] Hadi Sheriff Kolour ,Asadollah Shahbahrami“An
evaluation of license plate recognition algorithms”,
International Journal of digital information and wireless
communication (IJDIWC) 1(1):247-253.The society of
digital information&wireless communication (ISSN
2225-658X) 2011.
[12] Maha Mohamed Nabeel, Sherine Abd El-
Kader“Intelligent vehicle recognition based on wireless
sensor network “(IJCSI) international journal of
computer science vol.10,issue 4,no.2, July 2013.
[13] J.Zhou, S. Member, C. L. P. Chen, L. Chen, and W.
Zhao, “A user-customizable urban traffic information
collection method based on wireless sensor networks,”
IEEE Transaction on Intelligent Transportation
Systems.pp.1–10, Inpress2013.
[14] National Policing Improvement Agency,”Practice Advice
on the Management and Use Of Automatic Number Plate
Recognition”2009.
[15] Paolo Comelli M.N.G,Paolo Ferraggina F.Stabile,1995”
Optical recognition of motor vehicle license plate “IEEE
transcation on vehicular technology,44:790-799.