Use of Computer Vision Data for Detailed Driver Behavior Analysis and Trajectory Interpretation at Roundabouts

With recent important upgrades to North American intersection design guides, roundabouts are gaining popularity as a method of reducing road conflicts, streamlining flow, and curbing excessive speeding at busy intersections. The current design approach, however, makes use of mean spot speed measures and design criteria that do not take into account yielding behavior and acceleration and deceleration, all of which may be affected by regional driving culture and local roundabout design. This research paper introduces the methodology being developed for the detailed analysis of driving behavior, trajectory interpretation, and conflict measures in modern North American roundabouts, based on video data extracted by means of computer vision. The analysis explores the methods used to prepare microscopic speed maps, compiled speed profiles, lane-change counts, and gap time measures. The paper introduces and discusses the interpretation of trajectories at the scale of roundabout merge sections instead of looking at safety from the point of view of a roundabout as a unified system. The research found significant variation in distributions of speed across five case study roundabouts in the province of Quebec, Canada; this result may be explained by regional differences in design and road use. The paper also found aggressive gap times and uneven traffic flow as contributing factors to speed.