Automated Video Processing for Pedestrian-Vehicle Conflict Analysis

Pedestrian fatalities have risen in the United States over the past decade. On an individual corridor, however, it is difficult to determine whether crashes and fatalities are statistically significant or random occurrences. When considering mitigation efforts, transportation planners and engineers therefore need to accurately categorize pedestrian exposure and risk. Traditionally, risk and exposure were calculated by performing manual counts. Advancements in automated video processing, where objects are tracked from a recorded video, can categorize conflicts automatically. Using outputs from a developed tracking system, this paper defines a successful methodology to identify conflicts and calculate the post-encroachment time. This methodology can be applied to both intersection and non-intersection locations. Results from four sample sites support previous research that mid-block crossings occur more often when crosswalks are not nearby and the relationship between pedestrians and conflicts are not necessarily linear. Using this conflict identification methodology and automated video processing provides transportation planners and engineers with a better understanding of pedestrian risk at key locations.

• Record URL:
  https://annualmeeting.mytrb.org/OnlineProgram/Details/15772
• Supplemental Notes:
  • This paper was sponsored by TRB committee ACS20 Standing Committee on Safety Performance Analysis.
• Corporate Authors:

  Transportation Research Board

  ,    
• Authors:
  • Maddox, Spencer
  • Usher, Colin
  • Watkins, Kari
• Conference:
  • Transportation Research Board 100th Annual Meeting
  • Location: Washington DC, United States
  • Date: 2021-1-5 to 2021-1-29
• Date: 2021

Language

• English

Media Info

• Media Type: Digital/other
• Features: Figures; Photos; References; Tables;
• Pagination: 17p

Subject/Index Terms

• TRT Terms: Automation; Detection and identification systems; Image processing; Methodology; Midblock crossings; Pedestrian safety; Pedestrian vehicle crashes; Traffic conflicts; Traffic safety; Video
• Subject Areas: Data and Information Technology; Highways; Pedestrians and Bicyclists; Planning and Forecasting; Safety and Human Factors;

Filing Info

• Accession Number: 01764267
• Record Type: Publication
• Report/Paper Numbers: TRBAM-21-01654
• Files: TRIS, TRB, ATRI
• Created Date: Feb 4 2021 11:00AM