PRESENTATION ARCHIVE
OVERVIEW
Intelligent transportation systems (ITS) change our communities by improving the safety and convenience of people’s daily mobility. The system relies on multimodal traffic monitoring, that needs to provide reliable, efficient and detailed traffic information for traffic safety and planning. How to reliably and intelligently monitor intersection traffic with multimodal information is one of the most critical topics in intelligent transportation research.
In multimodal traffic monitoring, we gather traffic statistics for distinct transportation modes, such as pedestrians, cars and bicycles, in order to analyze and improve people’s daily mobility in terms of safety and convenience.
In this study, we use a high-resolution millimeter-wave (mmWave) radar sensor to obtain a relatively richer radar point cloud representation for a traffic monitoring scenario. Based on a new feature vector, we use the multivariate Gaussian mixture model (GMM) to do the radar point cloud segmentation, i.e. ‘pointwise’ classification, in an unsupervised learning environment.
In our experiment, we collected radar point clouds for pedestrians and cars, which also contained the inevitable clutter from the surroundings. The experimental results using GMM on the new feature vector demonstrated a good segmentation performance in terms of the intersection-over-union (IoU) metrics. The detailed methodology and validation metrics are presented and discussed.
KEY LEARNING OUTCOMES
- A mmWave radar based traffic monitoring system is presented.
- The radar can operate day and night under adverse weather conditions in traffic intersection.
- An unsupervised learning technique based on the multivariate Gaussian mixture model (GMM) is presented to segment the point cloud generated by radar.
THE RESEARCH
This webinar is based on a study funded by the National Institute for Transportation and Communities (NITC) and conducted at the University of Arizona. Read more about the research: Development Of Low-Cost Radar-Based Sensor For Multi-Modal Traffic Monitoring.
SPEAKER
Siyang Cao, University of Arizona
Siyang Cao joined the University of Arizona in 2015 as assistant professor of electrical and computer engineering, following a position as an automotive radar system engineer on algorithm, software and hardware development at Delphi. Cao is a graduate of The Ohio State University. His research focuses on the areas of radar signal processing, electronically scanned radar systems, radar imaging and machine learning with an emphasis on radar applications.
PROFESSIONAL DEVELOPMENT
This 60-minute webinar is eligible for 1 hour of professional development credit for AICP (see our provider summary). We provide an electronic attendance certificate for other types of certification maintenance.
LEARN MORE
Sign up for our newsletter to get updates on our events.
Image by Siyang Cao and Yao-Jan Wu, University of Arizona
This webinar is hosted by the Transportation Research and Education Center (TREC) at Portland State University. The research was funded by the Summit Foundation and the National Institute for Transportation and Communities (NITC), a program of TREC and one of five U.S. Department of Transportation national university transportation centers. The NITC program is a Portland State-led partnership with the University of Oregon, Oregon Institute of Technology, University of Utah and new partners University of Arizona and University of Texas at Arlington. We pursue our theme — improving mobility of people and goods to build strong communities — through research, education and technology transfer.