EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC 4.0

A Method for Localization and Dense Mapping in Indoor Dynamic Environments Based on an Improved ORB-SLAM2 Algorithm

Wanlei Liu, Yao Zhao

Journal of Computing and Electronic Information Management · 24.03.2026 · 20 (3) · 57-67

DOI: 10.54097/exr3jx13 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Computing and Electronic Information Management, online erschienen am 24.03.2026. Verfasst von Wanlei Liu und 1 weiteren Autor:innen. Die Arbeit nennt 28 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

To address the difficulty of ORB-SLAM2 in constructing dense maps in complex indoor dynamic environments, this paper proposes an improved method for localization and dense mapping in dynamic indoor scenes. The proposed approach integrates a semantic segmentation module based on DeepLabV3+ and introduces an additional semantic thread to obtain scene semantic information. In the tracking process, semantic segmentation results are combined with a dual depth-consistency checking strategy to identify and remove dynamic feature points while preserving reliable static features. Furthermore, the original keyframe selection mechanism is enhanced by incorporating relative pose variations between consecutive frames as an additional criterion, which helps reduce redundant keyframes and improve localization accuracy. A dense mapping thread is also introduced to fuse multi-frame point cloud data and generate a dense three-dimensional map of indoor environments. Experiments conducted on the TUM RGB-D Dataset demonstrate that the proposed method significantly improves trajectory accuracy. The average localization error is reduced by more than 95% in highly dynamic sequences and by 39.60% in low-dynamic sequences. The results show that the proposed approach achieves more accurate pose estimation and dense map reconstruction, providing effective support for mobile robot localization and navigation in indoor dynamic environments.

Zitieren

APA

Liu, W., & Zhao, Y. (2026). A Method for Localization and Dense Mapping in Indoor Dynamic Environments Based on an Improved ORB-SLAM2 Algorithm. Journal of Computing and Electronic Information Management, 20(3), 57-67. https://doi.org/10.54097/exr3jx13

Vancouver

Liu W, Zhao Y. A Method for Localization and Dense Mapping in Indoor Dynamic Environments Based on an Improved ORB-SLAM2 Algorithm. Journal of Computing and Electronic Information Management. 2026;20(3):57-67. doi:10.54097/exr3jx13

BibTeX
@article{Liu2026,
  author = {Liu, Wanlei and Zhao, Yao},
  title = {A Method for Localization and Dense Mapping in Indoor Dynamic Environments Based on an Improved ORB-SLAM2 Algorithm},
  journal = {Journal of Computing and Electronic Information Management},
  year = {2026},
  volume = {20},
  number = {3},
  pages = {57--67},
  doi = {10.54097/exr3jx13}
}
Kostenlos starten →