Highway work-zone safety is a major concern for government agencies, the legislature, and the traveling public.
Several work zone intelligent transportation systems (WZITS) have been developed as a safety countermeasure to
warn drivers of dangerous traffic conditions. Unfortunately, the effectiveness of a WZTIS is diminished if the
actual traffic flow conditions do not correspond with the sensor information leading to false warnings; these
confuse drivers and reduce the credibility of the system, which is often ignored. This can lead to situations where
drivers crash into work-zone areas because they are unprepared to stop. The national cost of crashes due to this was
estimated to be nearly $2.5 billion. Such “dangerous” traffic conditions are typically characterized by unpredictable
queue formations that propagate rapidly into higher speed traffic immediately upstream from the active work zone.
False positives or missed warnings could be reduced if the location of queue tails in addition to vehicle speeds in
proximity to the active work zone can be accurately detected. In this study, a low-cost rapidly deployable and
portable queue detection WZITS warning system is proposed. To demonstrate WZITS feasibility, a queue detection
algorithm was designed and tested using widely available, field proven, machine vision hardware that can be
integrated into the current portable system prototype, using video data collected in the field from the portable
device. The warning trigger generated by the algorithm can then be transmitted to a remote upstream location for
triggering roadside emergency warning devices (such as VMS, flashers, etc.).
Morris, Ted; Schwach, Jory A.; Michalopoulos, Panos G..
Low-Cost Portable Video-Based Queue Detection for Work- Zone Safety.
Retrieved from the University of Minnesota Digital Conservancy,
Content distributed via the University of Minnesota's Digital Conservancy may be subject to additional license and use restrictions applied by the depositor.