Analysis of Practical Methods for Counting Bicycling and Pedestrian Use of a Transportation Facility
Nikos Papanikolopoulos, Professor, Computer Science and Engineering
- Vassilios Morellas, Director, Computer Science and Engineering
Project Summary:Traffic counts of bicycles and pedestrians are essential in a variety of tasks, including the design of traffic control schemes, the design and modification of intersections, and the assessment of traffic measures such as stop signs and traffic signals, but manual counts are expensive and frequently inaccurate. This project examined available technologies for automated counting of bicycles and pedestrians in transportation facilities such as walk and bicycle bridges, urban bicycle routes, and bicycle trails, and highlighted the strengths and the pitfalls of each. The project?s goal was to create a matrix of methods and protocols that could be implemented by different municipalities and traffic engineers to evaluate the claims of the various vendors of traffic counters. This research also investigated counting systems based on computer vision technologies, in light of falling hardware costs and the fact that camera networks are becoming extremely common. A technical description of the various methods that were considered for vision-based object recognition is included in the report, as well as the reasoning behind such methods being overlooked for the project?s particular problem. After finalizing the software and hardware, five sites were picked for filming and about 10 hours of video was acquired in all. A portion of the video data was used for training and the remainder was used for testing the algorithm's accuracy. Count results and an interpretation of these results are also provided in this report. Finally, upon detailed analysis, the reasons for false counts and undercounting in certain cases were identified, as well as concerns and solutions for dealing with these issues.
- Start date: 10/2007
- Project Status: Completed
- Research Area: Transportation Safety and Traffic Flow
- Topics: Bicycling, Data and modeling, Pedestrian, Vision Systems