Scribd
Upload
55 views5 pages

Automatic & Robust Vehicle License Number Plate Recognition System (Electronics Project)

The document describes the development of an automatic vehicle license plate recognition system. It uses scan line evaluation and averaging to localize the license plate, then performs border removal and character mending to extract characters. A template matching approach recognizes characters. The system was tested on a variety of real images and achieved a 93% plate localization success rate and 96% character extraction success rate. Future work could enhance the system's ability to handle intricate license plate designs and non-uniform character heights.

Uploaded by

Vishal Agnihotri
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
55 views5 pages

Automatic & Robust Vehicle License Number Plate Recognition System (Electronics Project)

The document describes the development of an automatic vehicle license plate recognition system. It uses scan line evaluation and averaging to localize the license plate, then performs border removal and character mending to extract characters. A template matching approach recognizes characters. The system was tested on a variety of real images and achieved a 93% plate localization success rate and 96% character extraction success rate. Future work could enhance the system's ability to handle intricate license plate designs and non-uniform character heights.

Uploaded by

Vishal Agnihotri
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 5

ABSTRACT

An automatic and robust vehicle license plate recognition system has been
developed. The proposed method uses scan line evaluation and averaging method
to localize the number plate followed by a border removal mechanism combined
with character mending and approximation of character height to extract the
number plate characters.
Finally, a template matching approach is used to recognize the characters. A
Graphical User Interface has been created and the algorithm is experimented
successfully on a variety of real images, both single as well as double line plates.
The sample results obtained on testing with various images are also detailed.

LITERATURE REVIEW

Plate Localization

Plate localization is done to remove the unwanted background details, and thereby
focusing on to the essential details in the image. To detect the car plate, a
method by applying a top-hat filter to the whole image followed by a multiscale
region search has been described. Another approach has been proposed to detect
the vertical edges, to extract the license plate using sobel operators.

Character Extraction

Character extraction is done by segmenting the character portions from the


localized number plate. Noise contents that may also be present along with these
number plate characters, makes the job a tedious one. A method has been
proposed to remove the frame lines in the number plate followed by digit filtering.
A method on color image segmentation, region growing and clustering methods are
described.

Character Recognition

Character recognition is an essential and inevitable phase in such an application,


where the system is put up to deal with nebulous or distorted characters in the
license plate due to environmental hazards.

PROPOSED METHOD

Figure 1. Segmented region with three scan lines


As the image may have complex background details, number plate localization is
the central issue that demands great attention. Scan line evaluation and averaging
method is used here to accomplish this task. The image obtained from the sensor is
filtered and binarized and the set of connected components are segmented.
Following this step we measure the dimensions; height, width and area; of each of
these segmented regions. Only rectangular segmented regions are taken into
account.

Figure 3. (a) Original image, (b) Binarized image, (c) Localized plate region with information loss, (d) Localized plate
region after restoration

To extract the characters from the localized number plate, the image obtained
from the previous step is complemented. The contents in this image will either be
trivial noise components or characters to be identified. A border removal
mechanism followed by the approximation of character height is performed to
extract the characters.

Figure 5. (a) Localized number plate with distorted character within the box, (b) Mended character, (c) Segmentation
result of the distorted character (the three sections constitute to form the distorted character ‘5’)
But character breakage due to improper illumination can thwart the above
mentioned procedure of character extraction. So in order to surmount this problem
a character mending procedure is adopted as illustrated in figure 5. This process
strengthens the white pixels in the border removed number plate by turning its
adjacent left and top black pixels to white. This again is done by scanning through
the image for a white pixel.

EXPERIMENTAL RESULTS

Figure 6. (a) Original image, (b) Binarized image, (c) Localized number plate, (d) Extracted character set
Figure 7. (a) Original image, (b) Localized number plate, (c) Extracted character set

The concept used for coding the system does not limit it‘s workability to any
specific image format or size. The input to the system can either be a color or
grayscale image. Sample results obtained on working with jpeg and bmp formats
are presented here in figure 6 and figure 7. The system is designed in Matlab 7 for
recognizing the number plate characters of vehicles.

CONCLUSION AND FUTURE WORK

The paper explicates on license plate recognition emphasizing a scan line


evaluation method to localize the license plate and a border removal mechanism
combined with character mending & character height approximation to draw out
the characters from the license plate without much distortion. An effort has been
taken to make the proposed method work with variety of number plates of
different kind of vehicles of various countries.

The success rate on plate localization is about 93% and the success rate on
character extraction is about 96%. This can be taken as a feasible result when the
cost of implementation of the system is considered. This method works for those
images having a number plate with the characters inscribed in a single row or in
double row. The GUI provides a proper environment for the system to provide
promising results with white and black background number plates.

Depending upon the application for which the system is to be used, the distance
between the car plate & the camera unit is adjusted, to make it work properly.
Design of the number plate must be in such a manner that, their properties like
the color and reflective coatings etc, is to be within the bounds of the jurisdiction.
Intricate designs, that may be present in the foreground of the plate introduces a
higher dimension of complexity. This design complexity can be tackled if enough
information can be obtained about the foreground designs with respect to the
region where the system is deployed.

The vehicle may have any number of characters in its license plate and the system
does not readily depend upon the number of characters in the plate (-even, a prior
knowledge about the number of expected digits only enhances the workability-).
The effectiveness of the system is directly proportional to the readability of the
number plate. The system assumes that the characters inscribed in the license
plate are of uniform height. Finally, adequate amount of light energy and fine
weather conditions are preferred for better results.

You might also like