Probabilistic Framework for Automated Analysis of Exposure to Road Collisions

The advent of powerful sensing technologies, especially video sensors and computer vision techniques, has allowed for the collection of large quantities of detailed traffic data. These technologies allow further advancement toward completely automated systems for road safety analysis. This paper presents a comprehensive probabilistic framework for automated road safety analysis. Building on traffic conflict techniques and the concept of the safety hierarchy, it provides computational definitions of the probability of collision for road users involved in an interaction. It proposes new definitions for aggregated measures over time. This framework allows the interpretation of traffic from a safety perspective, by studying all interactions and their relationship to safety. New and more relevant exposure measures can be derived from this work, and traffic conflicts can be detected. A complete vision-based system is implemented to demonstrate the approach, providing experimental results on real-world video data.