Pedestrian Safety Diagnosis at Signalized Crosswalks Using Traffic Conflict Techniques

The objective of this study is the analysis of pedestrian safety at signalized crosswalks by using traffic conflict techniques. Video data was collected at a busy T-leg signalized intersection in Shanghai, China. Pedestrian-vehicle conflicts as well as pedestrian violations were automatically detected by analyzing the road users’ trajectories. It was found that the major cause of pedestrian-vehicle conflicts was right-turning vehicles that have a permitted right-turn phase. The right-turning vehicles involved in pedestrian-vehicle conflicts account for 64.81% of the total conflicts. The most serious pedestrian-vehicle conflicts occurred between pedestrians in crosswalks and vehicles from the crossing road traveling at high speed. It was found that more than half of the pedestrians were violation of signal timing or the crosswalk boundary: 37.81% were spatial violations and 18.46% were temporal violations. The results showed that conflicts associated with temporal violation behavior are much more numerous than spatial violation (48.41% and 17.06%).