Вы находитесь на странице: 1из 15

[Review/Dissertation] Seminar on

NUMBER PLATE RECOGNITION


for M. Tech. Degree
in the subject of Computer Science and Engineering
in the Faculty of Engineering and Technology
at
RASHTRASANT TUKADOJI MAHARAJ NAGPUR UNIVERSITY,
NAGPUR
SHREE K. LADDHA
STUDENT

Under the Guidance of


DR. A.J AGRAWAL
Associate Professor
Dept. of Comp. Sc. & Engg.

SHRI RAMDEOBABA COLLEGE OF ENGINEERING &


MANAGEMENT, NAGPUR
(An Autonomous College under RTM Nagpur University)

CONTENT
Introduction
Stages
Extraction
Segmentation
Character Recognition
Conclusion

2016

-- 2 --

INTRODUCTION
A vehicle registration plate is a metal or a plastic plate
attached to a motor vehicle for official identification purposes.
The number plates are placed on the front and back of the
vehicle.
This number is used for various applications.

2016

-- 3 --

STAGES

Fig. : Four Stages of an NPR System.


2016

-- 4 --

EXTRACTION
Extraction of the Number Plate from acquired image.
Input is the car image and output is the portion of the image.
The number plate has its own distinguishable features.
System processes the pixels having these features.

2016

-- 5 --

TECHNIQUES
Sobel operator
HSV colour model
Integral image

Fig. : A colour picture of


steam engine.

Fig. : Sobel operator applied


image.
2016

-- 6 --

HSV COLOR MODEL

2016

-- 7 --

INTEGRAL IMAGE

Algorithm for generating the sum of values in a rectangular subset


of a grid.
The value at any point (x,y) in the summed area table, i.e. sat(x,y),
is just the sum of all the pixels above and to the left of (x,y).

2016

-- 8 --

SEGMENTATION
Image segmentation is the process of partitioning a digital
image into multiple segments.
The goal of segmentation is to simplify and/or change the
representation of an image into something that is more
meaningful and easier to analyze.
Image segmentation is the process of dividing an image into
multiple parts.

2016

-- 9 --

SEGMENTATION

Fig. : Results of row segmentation by horizontal projection.

2016

-- 10

SEGMENTATION

Fig. : Character segmentation by vertical projection.

2016

-- 11

SEGMENTATION
Binary plate image is segmented to set of images which only
contain one license character.
These character are passed to the OCR module for
recognizing.
This procedure is done by using projections.
Horizontal projection is used to detect and segment rows.
Vertical projections are used to get images containing
individual characters.

2016

-- 12

CHARACTER RECOGNITION
The segmented images are passed to the OCR Module.
Hidden Markov Model is used.
Images used for training and images to deal with after well
trained are the same size of 5050 pixels.
Model is trained first using sets.

2016

-- 13

CONCLUSION
The system performs well on various types of VLP images.
In addition, it can deal with the cases of multiple plates in the
same image, or different types of vehicles such as motorbike
plates, car plates or truck plates.
Although there are some errors while dealing with bad
quality plates, the success rate is very good.

2016

-- 14

Thank You !!

2016

-- 15

Вам также может понравиться