Robust RGBD Visual Odometry Using Windowed Direct Bundle Adjustment and Slanted Support Plane

Publikation: Beitrag in FachzeitschriftArtikelForschungPeer-Review

Autoren

Forschungs-netzwerk anzeigen

Details

OriginalspracheEnglisch
Seiten (von - bis)350-357
Seitenumfang8
FachzeitschriftIEEE Robotics and Automation Letters
Jahrgang7
Ausgabenummer1
Frühes Online-Datum9 Nov. 2021
PublikationsstatusVeröffentlicht - 1 Jan. 2022

Abstract

This letter presents a robust, real-time visual odometry that efficiently exploits the available visual and geometry cues from RGB-D frames for both tracking and mapping. Together with a hybrid tracking algorithm based on a joint multi-objective formulation, we additionally incorporate the point-to-plane metrics into the photometric bundle adjustment (PBA) to constrain the iteration direction especially for those weakly-textured points. The relative pose constraints derived from optimized poses via PBA is then leveraged in combination with reprojection constraints retrieved from maintained feature map, to refine keyframe poses and feature locations. Moreover, the slanted support plane commonly used in multi-view stereo matching, is utilized for the adjustment of the semi-dense points to further enhance mapping accuracy that in turn benefits the front-end tracking. We extensively evaluate our algorithm on benchmark datasets, and those experimental results validate the advantage of our method in terms of overall tracking performance over other representative approaches.

ASJC Scopus Sachgebiete

Zitieren

Robust RGBD Visual Odometry Using Windowed Direct Bundle Adjustment and Slanted Support Plane. / Luo, Hang; Pape, Christian; Reithmeier, Eduard.
in: IEEE Robotics and Automation Letters, Jahrgang 7, Nr. 1, 01.01.2022, S. 350-357.

Publikation: Beitrag in FachzeitschriftArtikelForschungPeer-Review

Luo, H, Pape, C & Reithmeier, E 2022, 'Robust RGBD Visual Odometry Using Windowed Direct Bundle Adjustment and Slanted Support Plane', IEEE Robotics and Automation Letters, Jg. 7, Nr. 1, S. 350-357. https://doi.org/10.1109/LRA.2021.3126347
Luo H, Pape C, Reithmeier E. Robust RGBD Visual Odometry Using Windowed Direct Bundle Adjustment and Slanted Support Plane. IEEE Robotics and Automation Letters. 2022 Jan 1;7(1):350-357. Epub 2021 Nov 9. doi: 10.1109/LRA.2021.3126347
Luo, Hang ; Pape, Christian ; Reithmeier, Eduard. / Robust RGBD Visual Odometry Using Windowed Direct Bundle Adjustment and Slanted Support Plane. in: IEEE Robotics and Automation Letters. 2022 ; Jahrgang 7, Nr. 1. S. 350-357.
Download
@article{1b383aaae8d0483b91c0da1fa9def169,
title = "Robust RGBD Visual Odometry Using Windowed Direct Bundle Adjustment and Slanted Support Plane",
abstract = "This letter presents a robust, real-time visual odometry that efficiently exploits the available visual and geometry cues from RGB-D frames for both tracking and mapping. Together with a hybrid tracking algorithm based on a joint multi-objective formulation, we additionally incorporate the point-to-plane metrics into the photometric bundle adjustment (PBA) to constrain the iteration direction especially for those weakly-textured points. The relative pose constraints derived from optimized poses via PBA is then leveraged in combination with reprojection constraints retrieved from maintained feature map, to refine keyframe poses and feature locations. Moreover, the slanted support plane commonly used in multi-view stereo matching, is utilized for the adjustment of the semi-dense points to further enhance mapping accuracy that in turn benefits the front-end tracking. We extensively evaluate our algorithm on benchmark datasets, and those experimental results validate the advantage of our method in terms of overall tracking performance over other representative approaches. ",
keywords = "mapping, RGB-D images, Visual odometry",
author = "Hang Luo and Christian Pape and Eduard Reithmeier",
year = "2022",
month = jan,
day = "1",
doi = "10.1109/LRA.2021.3126347",
language = "English",
volume = "7",
pages = "350--357",
number = "1",

}

Download

TY - JOUR

T1 - Robust RGBD Visual Odometry Using Windowed Direct Bundle Adjustment and Slanted Support Plane

AU - Luo, Hang

AU - Pape, Christian

AU - Reithmeier, Eduard

PY - 2022/1/1

Y1 - 2022/1/1

N2 - This letter presents a robust, real-time visual odometry that efficiently exploits the available visual and geometry cues from RGB-D frames for both tracking and mapping. Together with a hybrid tracking algorithm based on a joint multi-objective formulation, we additionally incorporate the point-to-plane metrics into the photometric bundle adjustment (PBA) to constrain the iteration direction especially for those weakly-textured points. The relative pose constraints derived from optimized poses via PBA is then leveraged in combination with reprojection constraints retrieved from maintained feature map, to refine keyframe poses and feature locations. Moreover, the slanted support plane commonly used in multi-view stereo matching, is utilized for the adjustment of the semi-dense points to further enhance mapping accuracy that in turn benefits the front-end tracking. We extensively evaluate our algorithm on benchmark datasets, and those experimental results validate the advantage of our method in terms of overall tracking performance over other representative approaches.

AB - This letter presents a robust, real-time visual odometry that efficiently exploits the available visual and geometry cues from RGB-D frames for both tracking and mapping. Together with a hybrid tracking algorithm based on a joint multi-objective formulation, we additionally incorporate the point-to-plane metrics into the photometric bundle adjustment (PBA) to constrain the iteration direction especially for those weakly-textured points. The relative pose constraints derived from optimized poses via PBA is then leveraged in combination with reprojection constraints retrieved from maintained feature map, to refine keyframe poses and feature locations. Moreover, the slanted support plane commonly used in multi-view stereo matching, is utilized for the adjustment of the semi-dense points to further enhance mapping accuracy that in turn benefits the front-end tracking. We extensively evaluate our algorithm on benchmark datasets, and those experimental results validate the advantage of our method in terms of overall tracking performance over other representative approaches.

KW - mapping

KW - RGB-D images

KW - Visual odometry

UR - http://www.scopus.com/inward/record.url?scp=85120424457&partnerID=8YFLogxK

U2 - 10.1109/LRA.2021.3126347

DO - 10.1109/LRA.2021.3126347

M3 - Article

AN - SCOPUS:85120424457

VL - 7

SP - 350

EP - 357

JO - IEEE Robotics and Automation Letters

JF - IEEE Robotics and Automation Letters

SN - 2377-3766

IS - 1

ER -

Von denselben Autoren