Optical Character Recognition
on Vehicle Number Plate
Avinash Singh Bagri
2009MT50541
EEL 709
Dr Sumeet Agarwal
Overview
Takes image of the car and searches for the number plate in the image.
Once the probable number plate area is located it is given to OCR.
If OCR doesn’t recognize the characters from the image number plate area
is searched again from the image.
If characters are recognized then number plate search is terminated.
Steps Involved
Image division into small images
detecting probable number plate area
Recognizing number plate area
Parsing number plate to extract characters
Apply OCR to the parsed characters
Example
Original Image
Binarised image
Inverted binarised image
Extracting Number Plate
One piece of image that will be tested for number plate
Recognizing Plate
Search number plate in the broken pieces of vehicle image
Apply peak to valley to the candidate image pieces to further break the image piece
into possible character
Image piece with maximum peaks in candidate character is selected as the number
plate
Column signature of the number plate image Column signature of the another image piece
Parsing plate
Images of all characters
Recognition of Characters
Method of recognition of characters from an image containing these
characters is based on object recognition techniques used in Digital Image
Processing.
Two commonly used techniques
Template Matching using Correlation
Distance Measurement
Template Matching
Template matching using correlation
Based on performing correlation between segmented image
A character is required to be recognized and character template image
which is used for recognition
Correlation
Modified form of convolution
f(x,y): gray scale value at a specific element (x,y) in an image
f(x,y): image
g(x,y) character template
h(x,y) image after correlation.
Result of Correlation
Basic form of convolution
Result is an image, convolution of two matrices
The size of result matrix will be increased from input image matrices
Due to which we have to apply some thresh holding on resultant image
Normally value of thresh hold is little less than maximum value of resultant
image.
Limitations
Noise free image with uniform illumination required
Numbers must be displayed in one line on the number plate
Problem associated with template image is proper acquisition of template
image is required
References
http://www.sersc.org/journals/IJUNESST/vol6 no1/2.pdf
http://www.ele.uri.edu/~hansenj/projects/ele585/OCR/OCR.pdf
http://perun.pmf.uns.ac.rs/radovanovic/dmsem/completed/2006/OCR.pdf
http://www.nicomsoft.com/optical-character-recognition-ocr-how-it-works/
http://www.ancient-asia-journal.com/article/view/aa.06113/25
License Plate Number Recognition - New Heuristics and a Comparative Study of Classifiers
(2F79e4150656ca79aeb9.pdf)
Kwaśnicka H. and WawrzyniakLicense B.; Plate Localization and Recognition in Camera Pictures
Thank you

Optical Character Recognition