EUVIMEDEUVIMED entdecken
Crossref

AWC-LVI: Robust LiDAR-Visual-Inertial SLAM With Residual-Based Adaptive Weighting Controller in Complex Environments

Siyu He, Yingjing Qian, Fangyi Zhu, Fan Li, Zhaoxing Chen, Renmin Zhang

IEEE Sensors Letters · 2026-08 · 10 (8) · 5504504-5504504

DOI: 10.1109/lsens.2026.3707965 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Sensors Letters, 2026-08. Verfasst von Siyu He und 5 weiteren Autor:innen. Die Arbeit nennt 13 Literaturangaben.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Current light detection and ranging (LiDAR)-visual-inertial (LVI) simultaneous localization and mapping (SLAM) systems frequently encounter robustness bottlenecks in feature-scarce environments, due to the limitations of static measurement covariance assumptions. Such rigidity often leads to filter overconfidence in unreliable geometric constraints, suppressing critical state corrections from inertial priors. To address this issue, this letter proposes an adaptive tightly-coupled LVI–SLAM framework, where an adaptive weighting controller integrated with an extended error-state iterated Kalman filter architecture is constructed. By detecting geometric inconsistencies via real-time residual analysis, this mechanism dynamically rebalances the information matrix weight of the inertial measurement unit (IMU) constraint. This process effectively mitigates filter overconfidence in degenerate environments, ensuring an optimal fusion of LiDAR measurements and IMU motion priors. We conducted experiments on the NTU VIRAL dataset and the degraded dataset HILTI-22. Compared with the baseline FAST-LIVO2, our method has improved the average positioning accuracy by 8.7% in NTU, and is also leading on HILTI-22.

Zitieren

APA

He, S., Qian, Y., Zhu, F., Li, F., Chen, Z., & Zhang, R. (2026). AWC-LVI: Robust LiDAR-Visual-Inertial SLAM With Residual-Based Adaptive Weighting Controller in Complex Environments. IEEE Sensors Letters, 10(8), 5504504-5504504. https://doi.org/10.1109/lsens.2026.3707965

Vancouver

He S, Qian Y, Zhu F, Li F, Chen Z, Zhang R. AWC-LVI: Robust LiDAR-Visual-Inertial SLAM With Residual-Based Adaptive Weighting Controller in Complex Environments. IEEE Sensors Letters. 2026;10(8):5504504-5504504. doi:10.1109/lsens.2026.3707965

BibTeX
@article{He2026,
  author = {He, Siyu and Qian, Yingjing and Zhu, Fangyi and Li, Fan and Chen, Zhaoxing and Zhang, Renmin},
  title = {AWC-LVI: Robust LiDAR-Visual-Inertial SLAM With Residual-Based Adaptive Weighting Controller in Complex Environments},
  journal = {IEEE Sensors Letters},
  year = {2026},
  volume = {10},
  number = {8},
  pages = {5504504--5504504},
  doi = {10.1109/lsens.2026.3707965}
}
Kostenlos starten →