Traffic control by digital imaging cameras - Emerging Trends in Image Processing, Computer Vision, and Pattern Recognition - page 189

decades, researchers have deliberately worked on improving the control at traffic intersections
and traffic lights to reduce traffic jams and accidents [ 1 ] . The botleneck in traic problems is
related to the limited resources in the current infrastructure, such as road crossing or merging.
The traffic problems get worse with time, since the number of vehicles is increasing signiic-
antly.
The use of automated traffic control technologies is now wide spread throughout the world.
Worldwide, despite of variation in the nature of these applications, they have provided posit-
live road safety benefits. The first red-light camera was introduced in 1965 in the Netherlands.
This camera was based on using tubes stretched across the road to detect the violation and
subsequently trigger the camera [ 2 ]. New York's red-light camera program went into effect in
1993 [ 3 ].
The first digital camera system was introduced in Canberra in December 2000, and digital
cameras have increasingly replaced the old film cameras in other locations since then [ 4 ].
From the late 1990s, digital cameras began to be introduced, those cameras can be installed
with a network connection to transfer real-time live images to a central processing unit, for
that they have advantages over film cameras in speed monitoring. However, film-based sys-
tems may provide superior image quality in the variety of lighting conditions encountered on
roads. New film-based systems are still being used, but digital ones are providing greater pro-
iciency, lower maintenance and are now more popular.
2 Paper Overview
This paper discusses the production of a traffic control camera used to obtain red-light viol-
ation, license plate recognition, and speed detection of the vehicles. The proposed camera is
designed to be used at the botleneck of traic; intersections with traic lights. Since there are
several technologies used to obtain the same aim of this study, other characteristics were taken
into consideration to make it more atractive.
The major step in using such cameras is monitoring the traffic at the red light by capturing a
video. Then the video is processed by using image-processing techniques. The image-process-
ing code was developed using Matlab 7.7, whereby the program reads video file, converts it to
frames, and then by character segmentation, it can recognize the type of the violation and li-
license plate number. The main tasks of this camera include detection of the red-light violation.
Simply, the camera will check the color of the light, if it was red, it will compare between a
saved picture for the street in such red-light case (the street in front of the traffic light must be
empty) and the captured one. If there is any violation, the camera will capture a photo for the
car and perform plate recognition.
On the other hand, for the speed violation, the camera will measure the distance between
two points passed by the car and the time elapsed through this distance. Then, it will calculate
the speed of the car by dividing the distance over the time. Also, if there is any violation in
the speed, it will capture and perform plate recognition for the car. The system continuously
monitors the traffic signal and the camera is triggered by any vehicle entering the intersection.
Automatic number plate recognition can be used for purposes unrelated to enforcement of
traffic laws. In principle, any agency or person with access to data either from traffic cameras
or cameras installed for other purposes can track the movement of vehicles for any purpose.