investigating the influential factors in changing the likelihood of involving pedestrians in dangerous situations

Analyzing the pedestrian safety without using the accident data has become common in the recent years. investigating the pedestrian safety based on the conflict idea is becoming popular. therefore, this paper has investigated variables that cause interaction between vehicles and pedestrians for pedestrians, to be potentially dangerous and possibly critical situation. two measures were used in this study: time to collision (ttc) and postencroachment time (pet). first, an unsignalized intersection was chosen to launch this study. the site was filmed by camera for about 8 hours in two days. the different steps of this study were: identifying the conflict situations, tracking pedestrian and vehicles and then obtaining the pet, ttc, and the other data using matlab. probit models have been developed for analyzing the desired variables. there are 488 and 519 observations in ttc and pet models, respectively. in models with ttc being the dependent variable, the mean pedestrian and vehicle speed and the direction of pedestrian movement were some variables that cause an interaction to be potentially dangerous for pedestrians. furthermore, in models with pet being the dependent variable, mean vehicle and pedestrian speed, number of pedestrians, and the direction of vehicle movement were some factors that lead a conflict to possibly critical situation for pedestrians.