Modeling the Risk of Collision Between Pedestrians and Vehicles at Intersections

Safety is the main theme of this research as well as the main pillar of an optimal transportation system. Recent studies have shown that the cost of accidents and accidents is a huge part of the cost of traffic. Traffic interference techniques are supported as a preventive and complementary approach to road-based road safety analysis. However, the cost of field tracking of traffic conflicts with observer nature has limited the acceptance of these techniques. This research supports the use of a computer vision for automated and targeted traffic analysis. Video data in this cross-sectional study were collected in Babol city of Mazandaran province. The real coordinates of the users were extracted with calibration. After tracking road users, various time and space time measurements were used as a criterion for the risk of interference. Then using the artificial neural network, the risk of collision of pedestrians and cars was classified into 3 clusters; then, with the model presented in this study, new interference in a proportional risk cluster could be introduced using the hierarchical data analysis technique He anticipated using it in the field of safety studies before and after.