In this paper, we present a novel algorithm, namely
Curvefusion for integrating LiDAR scan matching with
stereo visual odometry. First, 6-DOF pose trajectories
are estimated by utilizing SOFT odometry, which is the
state of the art stereo visual odometry based on
feature selection and tracking, and the well-known ICP
scan matching algorithm, respectively. Second, a
deformation-based multi-sensor fusion method, namely
curvefusion is applied. The proposed fusion method
does not rely on a sensor model. As long as the
trajectories of the sensors to be fused are given, we
can obtain an optimized fusion trajectory, which
greatly improves the computational
efficiency. Experiments based on publicly available
KITTI data set show that the proposed method
outperforms or achieves similar performance compared
with the state-of-the-art odometry methods.
%0 Conference Paper
%1 ICCCR2021
%A Du, S.
%A Li, X.
%A Lauterbach, H. A.
%A Borrmann, D.
%A Nüchter, A.
%B Proceedings of the IEEE International Conference on Computer, Control and Robotics (ICCCR '21)
%D 2021
%K imported myown
%P 335--339
%R 10.1109/ICCCR49711.2021.9349385
%T Combining LiDAR Scan Matching with Stereo Visual Odometry
Using Curvefusion
%U https://robotik.informatik.uni-wuerzburg.de/telematics/download/icccr2021.pdf
%X In this paper, we present a novel algorithm, namely
Curvefusion for integrating LiDAR scan matching with
stereo visual odometry. First, 6-DOF pose trajectories
are estimated by utilizing SOFT odometry, which is the
state of the art stereo visual odometry based on
feature selection and tracking, and the well-known ICP
scan matching algorithm, respectively. Second, a
deformation-based multi-sensor fusion method, namely
curvefusion is applied. The proposed fusion method
does not rely on a sensor model. As long as the
trajectories of the sensors to be fused are given, we
can obtain an optimized fusion trajectory, which
greatly improves the computational
efficiency. Experiments based on publicly available
KITTI data set show that the proposed method
outperforms or achieves similar performance compared
with the state-of-the-art odometry methods.
@inproceedings{ICCCR2021,
abstract = {In this paper, we present a novel algorithm, namely
Curvefusion for integrating LiDAR scan matching with
stereo visual odometry. First, 6-DOF pose trajectories
are estimated by utilizing SOFT odometry, which is the
state of the art stereo visual odometry based on
feature selection and tracking, and the well-known ICP
scan matching algorithm, respectively. Second, a
deformation-based multi-sensor fusion method, namely
curvefusion is applied. The proposed fusion method
does not rely on a sensor model. As long as the
trajectories of the sensors to be fused are given, we
can obtain an optimized fusion trajectory, which
greatly improves the computational
efficiency. Experiments based on publicly available
KITTI data set show that the proposed method
outperforms or achieves similar performance compared
with the state-of-the-art odometry methods.},
added-at = {2021-03-14T14:35:44.000+0100},
author = {{Du}, S. and {Li}, X. and {Lauterbach}, H. A. and {Borrmann}, D. and {N{\"u}chter}, A.},
biburl = {https://www.bibsonomy.org/bibtex/2430d6e3bdd4e487e5cfd0ac4e861dc52/nuechter76},
booktitle = {Proceedings of the IEEE International Conference on Computer, Control and Robotics (ICCCR '21)},
doi = {10.1109/ICCCR49711.2021.9349385},
interhash = {71ba358f06b58ca7d5682aca4ef18f44},
intrahash = {430d6e3bdd4e487e5cfd0ac4e861dc52},
keywords = {imported myown},
pages = {335--339},
timestamp = {2021-06-08T10:44:42.000+0200},
title = {Combining LiDAR Scan Matching with Stereo Visual Odometry
Using Curvefusion},
url = {https://robotik.informatik.uni-wuerzburg.de/telematics/download/icccr2021.pdf},
year = 2021
}