الفهرس | Only 14 pages are availabe for public view |
Abstract Pedestrians, as vulnerable road users, are disproportionately affected by traffic collisions. However, assessing pedestrian safety in developing countries is challenging due to incomplete and inconsistent collision statistics. This study analyzes pedestrian safety in urban areas of Eastern Cairo, Egypt. The research introduces an innovative framework that combines macro-level safety analysis, utilizing crowdsourced data from the Streetguards website for hotspot location identification and micro-level safety analysis employing computer vision techniques with video data for automated pedestrian-vehicle conflict analysis. The macro-level analysis is executed using Geographical Information Systems (GIS). The hotspot identification process encompasses spatial analysis and spatial autocorrelation analysis. The outcomes indicate a general dispersion of incidents within the study area. Nevertheless, local spatial autocorrelation identifies specific locations on four major corridors as hotspots, with a confidence level of 99%. Based on the spatial analysis results, Abo Bakr El-Sedeek-Haroun El-Rashid crosswalk and El Hegaz location were selected for in-depth analysis. Three surrogate measures were used for conflict analysis, namely post-encroachment time (PET), road user speed, and the ratio of conflicting pedestrians. Additionally, a safety index (SI) was developed to assess the safety level of the study locations. The micro-level safety analysis utilizing video analytics to identify pedestrian-vehicle conflicts enhances the findings obtained in the initial stage of the research. Specifically, the analysis identifies Abo Bakr as a hotspot location with a lower level of safety compared to El Hegaz location. The framework presented in this study provides valuable insights for decision-makers to prioritize problematic locations associated with pedestrian safety and implement effective countermeasures to mitigate risks related to pedestrian crossings. |