Safety performance measures represent an increasingly more useful means for the evaluation of road safety conditions and allow the planners and the researchers to evaluate strategies apt to solve safety related problems. The main goal of this paper is to present a procedure for extracting vehicle tracking data and to use them in the estimation of safety performance in terms of a series of indicators representing interactions in real time between different pairs of vehicles belonging to the traffic stream. The context for this experiment is the measurement of safety performance at rural highways with respect to the potential for rear-end and head-on vehicle interactions. Individual vehicles are detected and tracked using a video image processing algorithm. The results of this study provide meaningful experimental indicators of potential safety problems at different rural highway locations subject to behavioral driver responses to different traffic conditions.
展开▼
