Loading [MathJax]/extensions/MathZoom.js
Application of semantic information module in LiDAR-based Simultaneous-Localization-and-Mapping algorithm | IEEE Journals & Magazine | IEEE Xplore

Application of semantic information module in LiDAR-based Simultaneous-Localization-and-Mapping algorithm


Abstract:

Autonomous driving and robotics require precise perception of the surrounding environment. Therefore, the simultaneous-localization-and-mapping algorithm, which is used f...Show More

Abstract:

Autonomous driving and robotics require precise perception of the surrounding environment. Therefore, the simultaneous-localization-and-mapping algorithm, which is used for accurate pose estimation and environmental perception, is of great importance. LiDAR, with its high measurement accuracy and ability to work around the clock, is widely used as the main sensor for simultaneous-localization-and-mapping algorithms. As the mass production technology of LiDAR becomes increasingly mature, various LiDAR based simultaneous-localization-and-mapping algorithms have been proposed one after another. In this paper, comparative experiments are conducted on the latest excellent algorithms, on this basis, a high-precision and robust LiDAR-based simultaneous-localization-and-mapping algorithm suitable for different scenarios is proposed. This algorithm combines the simultaneous-localization-and-mapping algorithm with the Scan Context semantic descriptor. Firstly, in the front end of the algorithm, the iterative Kalman filter is used for point cloud scan registration. In the back end, factor graph optimization is adopted to reduce global trajectory drift. Secondly, the algorithm performs loop closure detection from coarse to fine through Scan Context and the iterative closest point algorithm. Finally, the environmental map generated by the algorithm is managed by an incremental kd-tree. The evaluation experiments show that the algorithm proposed in this paper has significantly improved accuracy and robustness, which promotes the wide application of algorithms in different scenarios.
Published in: IEEE Access ( Early Access )
Page(s): 1 - 1
Date of Publication: 20 January 2025
Electronic ISSN: 2169-3536

Funding Agency: