Kristin Chen Discusses Computer Vision for Driver Research
How can computer vision be used to reduce video data in driver safety research? Kristin Chen, a Westat data scientist, will be describing such a process at the GASP! Government Advances in Statistical Programming! workshop. Ms. Chen’s work uses R, Python, and OpenCV tools to detect dashboard icons during a study of new car feature use. The Toyota Collaborative Safety Research Center in Ann Arbor, Michigan, is sponsoring the research.
The workshop is sponsored by the Washington Statistical Society. It will be held at the Bureau of Labor Statistics in Washington, DC, October 24-25, 2018.
Ms. Chen was successful using frame-extraction and feature-matching techniques to detect the emergence of driver-assist icons in dashboard videos captured during subject drives. This presentation (see Westat staff coauthors in bold below) describes the study background, the process, challenges, and performance metrics. These tools will be useful in future video-oriented driver safety studies.
Kristin Jiating Chen, Alexander Cates, Rick Huey, Marcelo Simas, James Jenness, Gonzalo Rivero. (Presentation). Computer Vision to Process Vehicle Dashboard Displays in Transportation Safety Research.