Visual-inertial motion priors for robust monocular SLAM

Usman Qayyum; Jonghyuk Kim

doi:10.1007/978-3-642-23232-9_62

Visual-inertial motion priors for robust monocular SLAM

Usman Qayyum^*, Jonghyuk Kim

^*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

Abstract

Monocular visual SLAM approaches are mostly constrained in their performance due to general motion model and availability of true scale information. We proposed an approach which improves the motion prediction step of visual SLAM and results in better estimation of map scale. The approach utilizes the short term accuracy of inertial velocity with visual orientation to estimate refined motion priors. These motion priors are fused with sparse number of 3D map features to constraint the positional drift of moving platform. Experimental results are presented on large scale outdoor environment, yielding robust performance and better observability of map scale by monocular SLAM.

Original language	English
Title of host publication	Towards Autonomous Robotic Systems - 12th Annual Conference, TAROS 2011, Proceedings
Pages	430-431
Number of pages	2
DOIs	https://doi.org/10.1007/978-3-642-23232-9_62
Publication status	Published - 2011
Event	12th Annual Conference on Towards Autonomous Robotic Systems, TAROS 2011 - Sheffield, United Kingdom Duration: 31 Aug 2011 → 2 Sept 2011

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	6856 LNAI
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	12th Annual Conference on Towards Autonomous Robotic Systems, TAROS 2011
Country/Territory	United Kingdom
City	Sheffield
Period	31/08/11 → 2/09/11

Access to Document

10.1007/978-3-642-23232-9_62

Cite this

Qayyum, U., & Kim, J. (2011). Visual-inertial motion priors for robust monocular SLAM. In Towards Autonomous Robotic Systems - 12th Annual Conference, TAROS 2011, Proceedings (pp. 430-431). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 6856 LNAI). https://doi.org/10.1007/978-3-642-23232-9_62

@inproceedings{06c9333796094ae2b96a079f7f254760,

title = "Visual-inertial motion priors for robust monocular SLAM",

abstract = "Monocular visual SLAM approaches are mostly constrained in their performance due to general motion model and availability of true scale information. We proposed an approach which improves the motion prediction step of visual SLAM and results in better estimation of map scale. The approach utilizes the short term accuracy of inertial velocity with visual orientation to estimate refined motion priors. These motion priors are fused with sparse number of 3D map features to constraint the positional drift of moving platform. Experimental results are presented on large scale outdoor environment, yielding robust performance and better observability of map scale by monocular SLAM.",

author = "Usman Qayyum and Jonghyuk Kim",

year = "2011",

doi = "10.1007/978-3-642-23232-9_62",

language = "English",

isbn = "9783642232312",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

pages = "430--431",

booktitle = "Towards Autonomous Robotic Systems - 12th Annual Conference, TAROS 2011, Proceedings",

note = "12th Annual Conference on Towards Autonomous Robotic Systems, TAROS 2011 ; Conference date: 31-08-2011 Through 02-09-2011",

}

Qayyum, U & Kim, J 2011, Visual-inertial motion priors for robust monocular SLAM. in Towards Autonomous Robotic Systems - 12th Annual Conference, TAROS 2011, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 6856 LNAI, pp. 430-431, 12th Annual Conference on Towards Autonomous Robotic Systems, TAROS 2011, Sheffield, United Kingdom, 31/08/11. https://doi.org/10.1007/978-3-642-23232-9_62

Visual-inertial motion priors for robust monocular SLAM. / Qayyum, Usman; Kim, Jonghyuk.
Towards Autonomous Robotic Systems - 12th Annual Conference, TAROS 2011, Proceedings. 2011. p. 430-431 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 6856 LNAI).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Visual-inertial motion priors for robust monocular SLAM

AU - Qayyum, Usman

AU - Kim, Jonghyuk

PY - 2011

Y1 - 2011

N2 - Monocular visual SLAM approaches are mostly constrained in their performance due to general motion model and availability of true scale information. We proposed an approach which improves the motion prediction step of visual SLAM and results in better estimation of map scale. The approach utilizes the short term accuracy of inertial velocity with visual orientation to estimate refined motion priors. These motion priors are fused with sparse number of 3D map features to constraint the positional drift of moving platform. Experimental results are presented on large scale outdoor environment, yielding robust performance and better observability of map scale by monocular SLAM.

AB - Monocular visual SLAM approaches are mostly constrained in their performance due to general motion model and availability of true scale information. We proposed an approach which improves the motion prediction step of visual SLAM and results in better estimation of map scale. The approach utilizes the short term accuracy of inertial velocity with visual orientation to estimate refined motion priors. These motion priors are fused with sparse number of 3D map features to constraint the positional drift of moving platform. Experimental results are presented on large scale outdoor environment, yielding robust performance and better observability of map scale by monocular SLAM.

UR - http://www.scopus.com/inward/record.url?scp=80052809390&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-23232-9_62

DO - 10.1007/978-3-642-23232-9_62

M3 - Conference contribution

SN - 9783642232312

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 430

EP - 431

BT - Towards Autonomous Robotic Systems - 12th Annual Conference, TAROS 2011, Proceedings

T2 - 12th Annual Conference on Towards Autonomous Robotic Systems, TAROS 2011

Y2 - 31 August 2011 through 2 September 2011

ER -

Visual-inertial motion priors for robust monocular SLAM

Abstract

Publication series

Conference

Access to Document

Other files and links

Fingerprint

Cite this