EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

ADM-SLAM: Accurate and Fast Dynamic Visual SLAM with Adaptive Feature Point Extraction, Deeplabv3pro, and Multi-View Geometry

Xiaotao Huang, Xingbin Chen, Ning Zhang, Hongjie He, Sang Feng

Sensors · 02.06.2024 · 24 (11) · 3578

DOI: 10.3390/s24113578 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Sensors, online erschienen am 02.06.2024. Verfasst von Xiaotao Huang und 4 weiteren Autor:innen. Beteiligte Einrichtung: School of Electromechanical Engineering, Guangdong University of Technology, Guangzhou 510006, China. Die Arbeit wurde bisher 14-mal zitiert (Crossref) und nennt 45 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: National Natural Science Foundation of China.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Visual Simultaneous Localization and Mapping (V-SLAM) plays a crucial role in the development of intelligent robotics and autonomous navigation systems. However, it still faces significant challenges in handling highly dynamic environments. The prevalent method currently used for dynamic object recognition in the environment is deep learning. However, models such as Yolov5 and Mask R-CNN require significant computational resources, which limits their potential in real-time applications due to hardware and time constraints. To overcome this limitation, this paper proposes ADM-SLAM, a visual SLAM system designed for dynamic environments that builds upon the ORB-SLAM2. This system integrates efficient adaptive feature point homogenization extraction, lightweight deep learning semantic segmentation based on an improved DeepLabv3, and multi-view geometric segmentation. It optimizes keyframe extraction, segments potential dynamic objects using contextual information with the semantic segmentation network, and detects the motion states of dynamic objects using multi-view geometric methods, thereby eliminating dynamic interference points. The results indicate that ADM-SLAM outperforms ORB-SLAM2 in dynamic environments, especially in high-dynamic scenes, where it achieves up to a 97% reduction in Absolute Trajectory Error (ATE). In various highly dynamic test sequences, ADM-SLAM outperforms DS-SLAM and DynaSLAM in terms of real-time performance and accuracy, proving its excellent adaptability.

Schlagwörter

Stichwörter der Autor:innen

V-SLAMdynamic environmentsfeature point extractionmulti-view geometrysemantic segmentation

Autor:innen und Institutionen

  • School of Electromechanical Engineering, Guangdong University of Technology, Guangzhou 510006, China.
  • Guangdong Productivity Promotion Center, Guangzhou 510075, China.

Förderung

  • National Natural Science Foundation of China

Zitieren

APA

Huang, X., Chen, X., Zhang, N., He, H., & Feng, S. (2024). ADM-SLAM: Accurate and Fast Dynamic Visual SLAM with Adaptive Feature Point Extraction, Deeplabv3pro, and Multi-View Geometry. Sensors, 24(11), 3578. https://doi.org/10.3390/s24113578

Vancouver

Huang X, Chen X, Zhang N, He H, Feng S. ADM-SLAM: Accurate and Fast Dynamic Visual SLAM with Adaptive Feature Point Extraction, Deeplabv3pro, and Multi-View Geometry. Sensors. 2024;24(11):3578. doi:10.3390/s24113578

BibTeX
@article{Huang2024,
  author = {Huang, Xiaotao and Chen, Xingbin and Zhang, Ning and He, Hongjie and Feng, Sang},
  title = {ADM-SLAM: Accurate and Fast Dynamic Visual SLAM with Adaptive Feature Point Extraction, Deeplabv3pro, and Multi-View Geometry},
  journal = {Sensors},
  year = {2024},
  volume = {24},
  number = {11},
  pages = {3578},
  doi = {10.3390/s24113578}
}
Kostenlos starten →