Abstract
High-definition mapping of 3D lane lines has been widely needed for the highway documentation and intelligent navigation of autonomous systems. A mobile mapping system (MMS) captures both accurate 3D LiDAR point clouds and high-resolution images of lane markings at highway driving speeds, providing an abundant data source for combined lane mapping. This paper aims to map lanes with an MMS. The main contributions of this paper include the following: (1) an intensity correction method was introduced to eliminate the reflectivity inconsistency of road-surface LiDAR points; (2) a self-adaptive thresholding method was developed to extract lane markings from their complicated surroundings; and (3) a LiDAR-guided textural saliency analysis of MMS images was proposed to improve the robustness of lane mapping. The proposed method was tested with a dataset acquired in Wuhan, Hubei, China, which contained straight roads, curved roads, and a roundabout with various pavement markings and a complex roadside environment. The experimental results achieved a recall of 96.4%, a precision of 97.6%, and an F-score of 97.0%, demonstrating that the proposed method has strong mapping ability for various urban roads.
Highlights
High-definition mapping of 3D lane lines is widely needed for various tasks [1,2], such as highway documentation, ego-vehicle localization with decimeter- or centimeter-level accuracy [3,4,5,6,7], behavior prediction of vehicles on the road [8,9], and safe route and evacuation plan generation [1,10,11] in autonomous driving systems
The 3D point clouds, images, and trajectory data that we need in this paper can be simultaneously obtained while the vehicle moves
Textural saliency analysis can be used in many cases, including the long and solid lane markings near the road boundary mentioned in Section 3.3.1, the dashed lane markings near the road boundaries or the center
Summary
High-definition mapping of 3D lane lines is widely needed for various tasks [1,2], such as highway documentation, ego-vehicle localization with decimeter- or centimeter-level accuracy [3,4,5,6,7], behavior prediction of vehicles on the road [8,9], and safe route and evacuation plan generation [1,10,11] in autonomous driving systems. In some GNSS-based studies [12,13,14,15,16], lane-level maps have been generated by driving along the centerline of the road (or lane) with a moving vehicle and analyzing the GNSS trajectory. Such methods are neither efficient nor accurate because each lane must be driven in multiple times and the path tends to deviate from the real centerline. Automated lane mapping could be improved by using more robust data sources
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
