01361669

Component

https://doi.org/10.3141/2241-10

http://scholar.google.com...aibi&publication_year=2011

It is plausible that road use behavior that leads to traffic violations can be a contributing factor to the failure mechanism that prompts a collision. The detection and the understanding of traffic violations, therefore, can be key components of a sound safety diagnosis. Recent advances in computer vision techniques have enabled the conduct of automated and large-scale analyses of various approaches to safety diagnoses, including traffic conflicts and violation analysis. This paper proposes and compares two approaches to detect vehicular spatial violations. The approaches are (a) k-means clustering and (b) pattern matching with use of the longest common subsequence (LCSS) similarity measure. The purpose of this study was to learn what constituted normal movement patterns and to interpret any discrepancy between these patterns and observed tracks as an indication of a traffic violation. Each of the approaches was applied to detect U-turn violations on an urban intersection in Kuwait City, Kuwait. Violation detection on the basis of LCSS generally was superior to detection with piecewise k-means clustering, especially when low false detection rates were desirable. This expected differential was the result of the performance of LCSS matching on observed road user positions, and not on an approximated model.

01361671

11-3532

English

pp 87-98

2011

Transportation Research Record: Journal of the Transportation Research Board

9780309167611

Print

Figures (8) ; References (28) ; Tables (1)

Intersections; Traffic violations; U turns; Video cameras

Automatic detection

Kuwait (Kuwait)

Highways; Safety and Human Factors; I83: Accidents and the Human Factor

TRIS, TRB, ATRI

Jan 30 2012 2:46PM

• Agent-Based Stochastic Modeling of Driver Decision at Onset of Yellow Light at Signalized Intersections
• Calibration of "Highway Safety Manual" Predictive Models for Oregon State Highways
• Evaluation of Texas Department of Transportation Safety Improvement Index as a Prioritization Tool
• Examination of Crash Variances Estimated by Poisson–Gamma and Conway–Maxwell–Poisson Models
• Highway Safety Metrics Implementation and Evaluation Using a Geographic Information System–Based Screening Tool
• Investigation of Effects of Underreporting Crash Data on Three Commonly Used Traffic Crash Severity Models: Multinomial Logit, Ordered Probit, and Mixed Logit
• Levels of Safety at Freeway Exits: Evaluations on the Basis of Individual Speed Difference
• Use of Empirical Bayesian Methods to Estimate Crash Modification Factors for Daytime Versus Nighttime Work Zones
• Use of Kriging Technique to Study Roundabout Performance
• Use of Light Detection and Ranging Data to Identify and Quantify Intersection Obstruction and Its Severity
• Workplace Simulator for Geometric Design of Rural Roads