| The detection of real-time traffic flow parameters plays a critical role in the intelligent transportation system. Parameter detection has many methods, and the video-based vehicle detection system has developed quickly in recently years. Since this approach has many outstanding advantages, such as wider-area detection and easily setting for system, it becomes a hot research area in intelligent transportation system.This paper focuses on developing a flexible and reliable system to detect the traffic flow parameters through image sequences. The vehicle detection is based on the roadway, which sets two virtual lines, like the inductive loop sensor, to detect their traffic flow parameters. The key idea of the system is converting the 2-dimensional digital image to 1-dimensional detection signal by virtual detection line preprocessing, which reduces computation load. The trait of this method is only processing little image area and avoiding vehicle tracking in 2-dimensional image space.In processing, improved background subtraction method saves vehicle information efficaciously. By comparing the common edge detection algorithms, the magniloquent Kirsch edge detection algorithm is selected to process the result of background subtraction. The result of edge detection is denoised by mathematical morphology method. Shadow edge restrained arithmetic and detection-line-based background updating method are presented. Shadow edge restrained arithmetic reduces vehicle shadow edge. Background updating precision is improved by presented background updating method. Last algorithms are designed for flow flux statistic, unit time traffic load, vehicle speed and roadway occupancy rate. Traffic flow parameter detection is implemented.The system of video-based traffic flow parameters detection using virtual line is implemented on a personal computer by VC++6.0. The traffic flow video files in different weather are experimented and the experimental results illustrate the detection system's accuracy.
|
PDF Full Text Request