Surround Mask Aiding GNSS/LiDAR SLAM for 3D Mapping in the Dense Urban Environment

Mengchi Ai, Yiran Luo, Naser El-Sheimy

Abstract:	Automatic driving with coupled Global Navigation Satellite System/ Inertial Navigation System (GNSS/INS) and light detection and ranging (LiDAR) sensors have been widely developed for many years, and successfully implemented in many stable urban environments. However, current approaches are still prone to errors, especially in dynamic and challenging environments. Specifically, an urban environment with plenty of moving objects and various building layouts and materials will bring unexpected non-rigid and abnormal features for LiDAR systems and multi-path for GNSS signals, respectively. To address the challenges in challenging environments, especially in urban scenarios, we propose an innovative scheme, namely surround mask, that explores the error sources from surrounding environments and then improves the performance of the integrated mapping system. Rather than applying complex post-processing to eliminate the accumulated error for each observing unit, the surround mask extracts a two-layer factor including NLOS detection and static objects detection to collectively compensate for the specific drawbacks of the LiDAR-based SLAM and the navigation system. Experimental results demonstrate that the proposed surround mask can detect the represent the error sources in the local coordinate and provide environment-awareness information for the integrated mapping system.
Published in:	Proceedings of the 35th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2022) September 19 - 23, 2022 Hyatt Regency Denver Denver, Colorado
Pages:	2011 - 2019
Cite this article:	Ai, Mengchi, Luo, Yiran, El-Sheimy, Naser, "Surround Mask Aiding GNSS/LiDAR SLAM for 3D Mapping in the Dense Urban Environment," Proceedings of the 35th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2022), Denver, Colorado, September 2022, pp. 2011-2019. https://doi.org/10.33012/2022.18550
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In