A Method to Distinguish between Congestion and Vehicle Accident Using Traffic Radar

Abstract

Individual radars measure range, radial speed, angle, reflectivity, and other parameters of multiple stationary and moving reflectors simultaneously. The measured signal is excellent for acquiring vehicle information on the road. Several studies have proposed identifying accident vehicles on the road by processing radar information. However, it is difficult to identify accident vehicles using a radar information in congested traffic since traffic radar information for accidents are almost the same as those for traffic jams. This study proposes a method to identifying accident vehicles based on an accident decision algorithm that distinguishes between congested and accident vehicles. Experimental results verified that the proposed method's detection accuracy exceeded 95%. We believe the proposed method will enhance future intelligent transportation systems.