01363301

Component

Automated pedestrian detection has the potential to enhance pedestrian safety, in a variety of applications. One application of recent interest is in extending signal timing, potentially as an alternative means of meeting new requirements for longer crossing times. However, the sorts of technologies used for automated detection are new and there is a need for field tests to demonstrate their reliability, in many cases shown to be insufficient, and to inform how they might be improved to be more useful tool for practitioners. The City of San Francisco installed an innovative system that uses video to identify the presence of late-crossing pedestrians to dynamically extend the protected crossing time by up to three seconds. Based on a set of observations completed in the summer of 2010, findings are presented on the device’s overall ability to detect pedestrians, to distinguish moving pedestrians from other pixel patterns in the video, and overall impact on the percent of pedestrians clearing the intersection on time as a result of the device. Overall, only a modest improvement is found; some of the factors affecting the performance of the device itself and implications for future research are discussed.

This paper was sponsored by TRB committee ANF10 Pedestrians

01362476

12-4638

English

15p

2012

Transportation Research Board 91st Annual Meeting

Digital/other

Figures; References; Tables

Detection and identification systems; Pedestrian actuated controllers; Pedestrian clearance interval (Traffic signals); Pedestrian safety; Time; Video imaging detectors

San Francisco (California)

Operations and Traffic Management; Pedestrians and Bicyclists; Safety and Human Factors; I73: Traffic Control; I83: Accidents and the Human Factor

Transportation Research Board Annual Meeting 2012 Paper #12-4638

TRIS, TRB

Feb 8 2012 5:25PM