Vehicle Number Plate Detection Using Sobel Edge Detection Techniques by MATLAB

Abstract

Detection of vehicle number plate is very interesting and also challenging topic for research.
Since identification of a particular vehicle can be possible through its number plate,
therefore each vehicle contains its own unique number plate. Characters are nothing but the
different shapes of lines or edges; hence, edge Detection is the most important step for
number plate extraction. Edges can be defined as the change in intensity of pixel, where each
pixel value is set on average value with its neighborhood pixel. Bright-pixel gets brighter and
dark-pixel gets darker hence characters are clearly visible which results into proper
segmentation and helps in number plate extraction. This paper presents an optimized method
based on Edge Detection Technique to identify the number plate of vehicle. As the number of
vehicles is increasing on the road, this calls for the need of traffic management, by this
method one can find whether the vehicle is registered or not. This also helps to maintain
smooth traffic movement and method used here simplifies the image characters, i.e. numbers
and alphabets, present in a number plate.