Title :
Visual road traffic monitoring and data collection
Author :
Dagless, E.L. ; Ali, A.T. ; Cruz, J.B.
Author_Institution :
Adv. Comput. Res. Centre, Bristol Univ., UK
Abstract :
The application of a parallel image processing system used for traffic monitoring is discussed. It is based on a multi-transputer system. The task of segmenting video images of a traffic scene is achieved by configuring a number of hardware modules to perform temporal smoothing, image differencing and thresholding. Feature extraction, motion prediction and brightness computation are assigned to a number of transputer framestores. Four lanes/entries can be monitored in real time using a total of six framestore units and three hardware video processing modules. The full system configuration is outlined and the hardware is described. Some results of monitoring motorway traffic and tracking vehicles in a four-arm junction are discussed.
Keywords :
traffic engineering computing; data collection; four-arm junction; hardware modules; hardware video processing modules; image differencing; image segmentation; motorway traffic; multi-transputer system; parallel image processing system; temporal smoothing; thresholding; traffic scene; transputer framestores; video images; visual road traffic monitoring; Brightness; Feature extraction; Hardware; Image processing; Image segmentation; Layout; Monitoring; Roads; Smoothing methods; Vehicles;
Conference_Titel :
Vehicle Navigation and Information Systems Conference, 1993., Proceedings of the IEEE-IEE
Conference_Location :
Ottawa, Ontario, Canada
Print_ISBN :
0-7803-1235-X
DOI :
10.1109/VNIS.1993.585604