Traffic control by digital imaging cameras - Emerging Trends in Image Processing, Computer Vision, and Pattern Recognition - page 198

(6)
After applying this part of the project the result of the plate number will appear in a text
document.
The templates that were used in finding the maximum correlation are provided in the sub-
sequent sections.
7 Performance analysis
The project was tested using a video for speed and red violations. For the plate position de-
termination and number recognition, a few images were tested.
7.1 Speed Violation
A video was tested for any speed violation; in the tested video, the car speed was 615 pixel/s
which violate the limit, while the other cars in the street had a speed in the range of
(500-580)pixel/s.
There is a ratio for transformation from pixel to meter each 1 pixel equals 0.000264583 m,
but this is not quit accurate since the zoom ratio affect number and size of the pixel.
7.2 Red Violation
For the same video, the car has violated the red light, and the project code was able to detect
it.
7.3 Plate Position Determination
This part from the project has worked efficiently with some errors.
The proposed design was tested at 25 plates; the accuracy percentage was determined by
the following:
(7)
Note : Any addition number appears in the result count as an error, Table 1 shows the results
and the percentage of error.