Home // GEOProcessing 2022, The Fourteenth International Conference on Advanced Geographic Information Systems, Applications, and Services // View article

Towards Automatic Inference of Layouts of Traffic Intersections for Smart Cities

Authors:
Julien A. Vijverberg
Bart J. Beers
Peter H. N. de With

Keywords: machine vision; terrain mapping; road transportation

Abstract:
Intersection-topology descriptions can help to improve traffic flow, but currently require significant manual creation effort. This paper describes ongoing work on a combination of algorithms to extract the drive lines across intersections from images and point clouds, captured by a ground-based vehicle. The extraction is based on paint striping and edge-of-road line features, which are clustered in subsequent stages into lane separators for all legs of the intersection. The GEO recall with paint striping is~0.29, and significantly contributes to successfully inferring the junction. However, this score decreases when edge-of-road features are added which is counter-intuitive. We indicate ways to combat this problem and further improve our results.

Pages: 43 to 46

Copyright: Copyright (c) IARIA, 2022

Publication date: June 26, 2022

Published in: conference

ISSN: 2308-393X

ISBN: 978-1-61208-983-6

Location: Porto, Portugal

Dates: from June 26, 2022 to June 30, 2022