Vehicle Registration
Plate Recognition
System
BY: SHAILENDRA GUPTA
Contents
 Introduction and Objective of the project
 Methodology
 Pretreatment of the Image
Introduction and Objective
 For an important method of car control, we require an system for
vehicle plate recognition.
 This system will provide for monitoring vehicles for law enforcement
and security purposes, and keeping track of vehicles by successful
recognition of plate number.
 May be used in places such as border control and Automatic Toll
tax collection
Methodology
 The system for plate recognition consists of Three main stages:
 (1) Preprocessing
 (2) License plate extraction
 (3) Template Matching and Character recognition
Preprocessing of the Image
 Graying of Original Image
 As problem lies with the data size of original plate image is the larger
one, the calculated amount of direct processing these colors is
extremely large, so transform the original image into 256-level gray
image, convenient for computer processing. Gray-scale image only
have brightness information instead of color information
 Resizing the image
 To fit the window we have to resize the image without changing
aspect ratio.
Histogram Equalization:
Removing Small Objects via
erosion
Licence Plate Extraction
 Sobel edge detector
 The next step involves highlighting the horizontal edges and removing them for
better plate area.
 This is done by convolving the sobel operated image which is a polynomial
multiplication, which results in bright horizontal lines.
 Now deleting these horizontal lines as they may be the edges of the licence plate.
 We will use erosion function with structuring element a line having angle 0 degrees
for horizontal direction and subtracting from the convoluted image of sobel
operated image.
 After these steps cropping function in matlab is used to extract the licence plate
area.
Template matching and character
recognition
 After extracting the plate area template matching is done to
successfully recognise the Registration number of a car.
Thank You

Vehicle registration plate recognition system

  • 1.
  • 2.
    Contents  Introduction andObjective of the project  Methodology  Pretreatment of the Image
  • 3.
    Introduction and Objective For an important method of car control, we require an system for vehicle plate recognition.  This system will provide for monitoring vehicles for law enforcement and security purposes, and keeping track of vehicles by successful recognition of plate number.  May be used in places such as border control and Automatic Toll tax collection
  • 4.
    Methodology  The systemfor plate recognition consists of Three main stages:  (1) Preprocessing  (2) License plate extraction  (3) Template Matching and Character recognition
  • 5.
    Preprocessing of theImage  Graying of Original Image  As problem lies with the data size of original plate image is the larger one, the calculated amount of direct processing these colors is extremely large, so transform the original image into 256-level gray image, convenient for computer processing. Gray-scale image only have brightness information instead of color information  Resizing the image  To fit the window we have to resize the image without changing aspect ratio.
  • 6.
  • 7.
  • 8.
    Licence Plate Extraction Sobel edge detector
  • 9.
     The nextstep involves highlighting the horizontal edges and removing them for better plate area.  This is done by convolving the sobel operated image which is a polynomial multiplication, which results in bright horizontal lines.  Now deleting these horizontal lines as they may be the edges of the licence plate.  We will use erosion function with structuring element a line having angle 0 degrees for horizontal direction and subtracting from the convoluted image of sobel operated image.  After these steps cropping function in matlab is used to extract the licence plate area.
  • 10.
    Template matching andcharacter recognition  After extracting the plate area template matching is done to successfully recognise the Registration number of a car.
  • 11.