Journal
IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING
Volume 12, Issue 5, Pages 1572-1586Publisher
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/JSTARS.2019.2904514
Keywords
Cubic spline; driving line; equidistant line; high-definition (HD) map; mobile laser scanning (MLS); point clouds; road marking; road surface
Categories
Funding
- National Natural Science Foundation of China [41471379]
- Natural Sciences and Engineering Research Council of Canada [50503-10284]
Ask authors/readers for more resources
This paper presents the development of a semiautomated driving line generation method using point clouds acquired by a mobile laser scanning system. Horizontally curved driving lines are a critical component for high-definition maps that are required by autonomous vehicles. The proposed method consists of three steps: Road surface extraction, road marking extraction, and driving line generation. First, the points covering road surfaces are extracted using the curb-based road surface extraction algorithms depending on both the elevation and slope differences. Then, road markings are identified and extracted according to a variety of algorithms consisting of georeferenced intensity imagery generation, multithreshold road marking extraction, and statistical outlier removal. Finally, the conditional Euclidean clustering algorithm is employed, followed by the cubic spline curve-fitting algorithm and equidistant line-based driving line generation algorithms for horizontally curved driving line generation. Our method is evaluated by six MLS point cloud datasets collected from various types of horizontally curved road corridors. Quantitative evaluations demonstrate that the proposed road marking extraction algorithm achieves an average recall, precision, and F1-score of 90.79%, 92.94%, and 91.85%, respectively. The generated driving lines are assessed by overlaying them on the manually interpreted reference buffers from 4-cm resolution unmanned aerial vehicle orthoimagery, and a 15 cm level navigation and localization accuracy is achieved.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available