Proceedings of Engineering and Technology Innovation (Jul 2019)

Vision Technology Based Traffic Safety Analysis Using Signal Data

  • Wonho Suh,
  • Jung Ryul Kim,
  • Joongmin Cho

Journal volume & issue
Vol. 13
pp. 20 – 25


Read online

Recent vision technology allows traffic engineers to analyze traffic safety based on image processing applications. Vehicle trajectory data including vehicle position and speed are extracted and converted into traffic conflicts related variables. These variables are analyzed to find correlations with crash data and geometric design variables and included in the statistical analysis process to evaluate the safety of intersections. In this paper, signal timing information data is included in the analysis process to investigate if there are any correlations with traffic conflict data. For example, traffic conflicts happen more frequently in a certain movement at certain signal time phase. The goal of this paper is to develop a method for vision technology-based traffic safety analysis process using traffic signal data to identify crash prone movement and signal to time at a given intersection. The proposed technique is demonstrated in real-world video data collected in an intersection. This paper is expected to provide more insight and technique in traffic safety evaluation.