Traffic Queue Detection Using Image-Processing

by A. Rourke, Univ of Newcastle Upon Tyne, United Kingdom,
M. G. H. Bell, Univ of Newcastle Upon Tyne, United Kingdom,

Document Type: Proceeding Paper

Part of: Applications of Advanced Technologies in Transportation Engineering


A method of queue/congestion detection is described using low cost image processing and computing hardware. It is capable of processing data from a length of roadway within a few seconds, providing distinct advantages over more conventional methods. It is seen thatthe current system is capable of providing a response to queueing or congested traffic within 4 to 8 seconds of the queueing occuring. A measure of the extent of the congestion can also be obtained along with the location of the rear of the queue. Further work is required to improve the speed of the algorithm and to allow processing of calibrated scenes. An increase in execution speed can be achieved by employing a digital signal processing (DSP) co-processor to perform the Fast Fourier Transform (FFT) analysis. This will allow multiple windows to be analysed within the acceptable time-scale. Scene calibration is required to account for minor perspective distortion effects and to enable queue parameters, such as queue length and location of rear end, to become physically meaningful.

Subject Headings: Traffic congestion | Computer vision and image processing | Traffic analysis | Traffic surveys | Traffic management | Calibration | Data analysis | Signal processing

Services: Buy this book/Buy this article


Return to search