EUVIMEDEUVIMED entdecken
Crossref

MUD-SLAM: Measurement Uncertainty-Driven Loop Closing Method for Robust LiDAR–IMU SLAM in Complex Environments

Mohan Jin, Nan Zang, Jianhua Cheng, Runze Bai, Sicheng Chen, Peng Li

IEEE Transactions on Instrumentation and Measurement · 2026 · 75 · 8508716-8508716

DOI: 10.1109/tim.2026.3695048 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Transactions on Instrumentation and Measurement, 2026. Verfasst von Mohan Jin und 5 weiteren Autor:innen. Die Arbeit nennt 38 Literaturangaben. Gefördert von: National Key Research and Development Program of China, National Natural Science Foundation of China, Heilongjiang Province Youth Science and Technology Talent.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Loop closure detection (LCD) and global consistency constraint (GCC) constitute critical components in light detection and ranging (LiDAR)–inertial measurement unit (IMU) simultaneous localization and mapping (SLAM) systems. Neglecting the spatial uncertainty propagating from measurement uncertainty into point clouds can significantly impair both LCD and GCC robustness, ultimately degrading pose estimation accuracy in complex environments. To address this limitation, we propose measurement uncertainty-driven SLAM (MUD-SLAM), a probabilistic constraint graph framework that explicitly models measurement uncertainty distributions for covariance-aware loop closing. Specifically, using plane-fitting as a reference, we first establish a polynomial regression model for range measurement error distributions correlated with incidence angles. This model enables rigorous derivation of point cloud spatial uncertainty through error propagation. Hereby, a reliable LCD (RLCD) method is proposed, which integrates a robust K–D tree algorithm based on probabilistic geometric–semantic fusion for filtering loop candidate frames and an IMU-assisted geometric verification strategy to actively suppress false registration in feature degradation scenarios. Furthermore, an adaptive GCC (AGCC) model is constructed by the covariance-aware weighted strategy to achieve resilient optimization of loop closure. Comprehensive evaluations on complex and deteriorated environment datasets demonstrate MUD-SLAM’s superior accuracy and timeliness in bidirectional loops. Especially in adverse weather, about 21.61%, 15.71%, and 6.72% improvements of 3-D positioning accuracy are achieved in MUD-SLAM compared to LOAM, LEGO-LOAM, and LIO-SAM. MUD-SLAM provides new insights into enhancing the robustness against the degradations of the environment and sensors.

Förderung

  • National Key Research and Development Program of China
  • National Natural Science Foundation of China
  • Heilongjiang Province Youth Science and Technology Talent
  • Heilongjiang Provincial Key Research and Development Program
  • China Postdoctoral Science Foundation

Zitieren

APA

Jin, M., Zang, N., Cheng, J., Bai, R., Chen, S., & Li, P. (2026). MUD-SLAM: Measurement Uncertainty-Driven Loop Closing Method for Robust LiDAR–IMU SLAM in Complex Environments. IEEE Transactions on Instrumentation and Measurement, 75, 8508716-8508716. https://doi.org/10.1109/tim.2026.3695048

Vancouver

Jin M, Zang N, Cheng J, Bai R, Chen S, Li P. MUD-SLAM: Measurement Uncertainty-Driven Loop Closing Method for Robust LiDAR–IMU SLAM in Complex Environments. IEEE Transactions on Instrumentation and Measurement. 2026;75:8508716-8508716. doi:10.1109/tim.2026.3695048

BibTeX
@article{Jin2026,
  author = {Jin, Mohan and Zang, Nan and Cheng, Jianhua and Bai, Runze and Chen, Sicheng and Li, Peng},
  title = {MUD-SLAM: Measurement Uncertainty-Driven Loop Closing Method for Robust LiDAR–IMU SLAM in Complex Environments},
  journal = {IEEE Transactions on Instrumentation and Measurement},
  year = {2026},
  volume = {75},
  pages = {8508716--8508716},
  doi = {10.1109/tim.2026.3695048}
}
Kostenlos starten →