Details
Original language | English |
---|---|
Article number | 9392262 |
Pages (from-to) | 4899-4906 |
Number of pages | 8 |
Journal | IEEE Robotics and Automation Letters |
Volume | 6 |
Issue number | 3 |
Publication status | Published - 31 Mar 2021 |
Abstract
This letter presents a hybrid framework, both in front-end and back-end, for monocular simultaneous localization and mapping (SLAM), capable of utilizing the robustness of feature matching and the accuracy of direct alignment. In the front-end, the feature-based method is first used for coarse pose estimation that is subsequently taken by the direct alignment module as initialization for further refinement. In the back-end, a double window structure is constructed based on the maintained semi-dense map and the sparse feature map, of which the states are optimized via a multi-layer optimization scheme based on the reprojection constraints and the relative pose constraints. Our evaluation on several public datasets demonstrates that this hybrid design retains the superior resilience to scene variations of salient features, and achieves better tracking accuracy due to the integration of the direct modules, leading to a comparable performance with the state-of-the-arts.
Keywords
- double window optimization, mapping, SLAM
ASJC Scopus subject areas
- Engineering(all)
- Control and Systems Engineering
- Engineering(all)
- Biomedical Engineering
- Computer Science(all)
- Human-Computer Interaction
- Engineering(all)
- Mechanical Engineering
- Computer Science(all)
- Computer Vision and Pattern Recognition
- Computer Science(all)
- Computer Science Applications
- Mathematics(all)
- Control and Optimization
- Computer Science(all)
- Artificial Intelligence
Cite this
- Standard
- Harvard
- Apa
- Vancouver
- BibTeX
- RIS
In: IEEE Robotics and Automation Letters, Vol. 6, No. 3, 9392262, 31.03.2021, p. 4899-4906.
Research output: Contribution to journal › Article › Research › peer review
}
TY - JOUR
T1 - Hybrid Monocular SLAM Using Double Window Optimization
AU - Luo, Hang
AU - Pape, Christian
AU - Reithmeier, Eduard
N1 - Funding Information: Manuscript received December 7, 2020; accepted March 21, 2021. Date of publication March 31, 2021; date of current version April 20, 2021. This letter was recommended for publication by Associate Editor Y. Latif and Editor J. Civera upon evaluation of the reviewers’ comments. This work was supported by China Scholarship Council. (Corresponding author: Hang Luo.) The authors are with the Institute of Measurement and Automatic Control, Faculty of Mechanical Engineering, Leibniz University Hannover, 30167 Hanover, Germany (e-mail: luo@imr.uni-hannover.de; Christian.Pape@imr.uni-hannover.de; reithmeier@imr.uni-hannover.de). Digital Object Identifier 10.1109/LRA.2021.3070298
PY - 2021/3/31
Y1 - 2021/3/31
N2 - This letter presents a hybrid framework, both in front-end and back-end, for monocular simultaneous localization and mapping (SLAM), capable of utilizing the robustness of feature matching and the accuracy of direct alignment. In the front-end, the feature-based method is first used for coarse pose estimation that is subsequently taken by the direct alignment module as initialization for further refinement. In the back-end, a double window structure is constructed based on the maintained semi-dense map and the sparse feature map, of which the states are optimized via a multi-layer optimization scheme based on the reprojection constraints and the relative pose constraints. Our evaluation on several public datasets demonstrates that this hybrid design retains the superior resilience to scene variations of salient features, and achieves better tracking accuracy due to the integration of the direct modules, leading to a comparable performance with the state-of-the-arts.
AB - This letter presents a hybrid framework, both in front-end and back-end, for monocular simultaneous localization and mapping (SLAM), capable of utilizing the robustness of feature matching and the accuracy of direct alignment. In the front-end, the feature-based method is first used for coarse pose estimation that is subsequently taken by the direct alignment module as initialization for further refinement. In the back-end, a double window structure is constructed based on the maintained semi-dense map and the sparse feature map, of which the states are optimized via a multi-layer optimization scheme based on the reprojection constraints and the relative pose constraints. Our evaluation on several public datasets demonstrates that this hybrid design retains the superior resilience to scene variations of salient features, and achieves better tracking accuracy due to the integration of the direct modules, leading to a comparable performance with the state-of-the-arts.
KW - double window optimization
KW - mapping
KW - SLAM
UR - http://www.scopus.com/inward/record.url?scp=85103793475&partnerID=8YFLogxK
U2 - 10.1109/LRA.2021.3070298
DO - 10.1109/LRA.2021.3070298
M3 - Article
AN - SCOPUS:85103793475
VL - 6
SP - 4899
EP - 4906
JO - IEEE Robotics and Automation Letters
JF - IEEE Robotics and Automation Letters
SN - 2377-3766
IS - 3
M1 - 9392262
ER -