Consistency analysis and improvement for single-camera localization

Joel A. Hesch; Stergios Roumeliotis

doi:10.1109/CVPRW.2012.6239190

Consistency analysis and improvement for single-camera localization

Joel A. Hesch, Stergios Roumeliotis

Computer Science and Engineering

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

11 Scopus citations

Abstract

In this paper, we study the problem of estimator inconsistency in single-camera simultaneous localization and mapping (MonoSLAM) from a standpoint of system observability. Specifically, we postulate that a leading cause of inconsistency is the gain of spurious information along unobservable directions, resulting in smaller uncertainties, larger estimation errors, and divergence. Moreover, we introduce an Observability-Constrained MonoSLAM (OC-MonoSLAM) approach, which explicitly enforces the unobservable directions of the system, hence preventing spurious information gain and reducing inconsistency. Our analysis, along with the proposed method for reducing inconsistency, are validated with simulation trials and real-world experimentation.

Original language	English (US)
Title of host publication	2012 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2012
Pages	15-22
Number of pages	8
DOIs	https://doi.org/10.1109/CVPRW.2012.6239190
State	Published - Aug 20 2012
Event	2012 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2012 - Providence, RI, United States Duration: Jun 16 2012 → Jun 21 2012

Publication series

Name	IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops
ISSN (Print)	2160-7508
ISSN (Electronic)	2160-7516

Other

Other	2012 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2012
Country/Territory	United States
City	Providence, RI
Period	6/16/12 → 6/21/12

Publisher link

10.1109/CVPRW.2012.6239190

Find It

Find It at U of M Libraries

Cite this

Hesch, J. A., & Roumeliotis, S. (2012). Consistency analysis and improvement for single-camera localization. In 2012 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2012 (pp. 15-22). Article 6239190 (IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops). https://doi.org/10.1109/CVPRW.2012.6239190

Consistency analysis and improvement for single-camera localization. / Hesch, Joel A.; Roumeliotis, Stergios.
2012 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2012. 2012. p. 15-22 6239190 (IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops).

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

Hesch, JA & Roumeliotis, S 2012, Consistency analysis and improvement for single-camera localization. in 2012 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2012., 6239190, IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, pp. 15-22, 2012 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2012, Providence, RI, United States, 6/16/12. https://doi.org/10.1109/CVPRW.2012.6239190

@inproceedings{d0fb4b5cbb724a66ae8102f492ea97d8,

title = "Consistency analysis and improvement for single-camera localization",

abstract = "In this paper, we study the problem of estimator inconsistency in single-camera simultaneous localization and mapping (MonoSLAM) from a standpoint of system observability. Specifically, we postulate that a leading cause of inconsistency is the gain of spurious information along unobservable directions, resulting in smaller uncertainties, larger estimation errors, and divergence. Moreover, we introduce an Observability-Constrained MonoSLAM (OC-MonoSLAM) approach, which explicitly enforces the unobservable directions of the system, hence preventing spurious information gain and reducing inconsistency. Our analysis, along with the proposed method for reducing inconsistency, are validated with simulation trials and real-world experimentation.",

author = "Hesch, {Joel A.} and Stergios Roumeliotis",

year = "2012",

month = aug,

day = "20",

doi = "10.1109/CVPRW.2012.6239190",

language = "English (US)",

isbn = "9781467316118",

series = "IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops",

pages = "15--22",

booktitle = "2012 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2012",

note = "2012 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2012 ; Conference date: 16-06-2012 Through 21-06-2012",

}

TY - GEN

T1 - Consistency analysis and improvement for single-camera localization

AU - Hesch, Joel A.

AU - Roumeliotis, Stergios

PY - 2012/8/20

Y1 - 2012/8/20

N2 - In this paper, we study the problem of estimator inconsistency in single-camera simultaneous localization and mapping (MonoSLAM) from a standpoint of system observability. Specifically, we postulate that a leading cause of inconsistency is the gain of spurious information along unobservable directions, resulting in smaller uncertainties, larger estimation errors, and divergence. Moreover, we introduce an Observability-Constrained MonoSLAM (OC-MonoSLAM) approach, which explicitly enforces the unobservable directions of the system, hence preventing spurious information gain and reducing inconsistency. Our analysis, along with the proposed method for reducing inconsistency, are validated with simulation trials and real-world experimentation.

AB - In this paper, we study the problem of estimator inconsistency in single-camera simultaneous localization and mapping (MonoSLAM) from a standpoint of system observability. Specifically, we postulate that a leading cause of inconsistency is the gain of spurious information along unobservable directions, resulting in smaller uncertainties, larger estimation errors, and divergence. Moreover, we introduce an Observability-Constrained MonoSLAM (OC-MonoSLAM) approach, which explicitly enforces the unobservable directions of the system, hence preventing spurious information gain and reducing inconsistency. Our analysis, along with the proposed method for reducing inconsistency, are validated with simulation trials and real-world experimentation.

UR - http://www.scopus.com/inward/record.url?scp=84865044286&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84865044286&partnerID=8YFLogxK

U2 - 10.1109/CVPRW.2012.6239190

DO - 10.1109/CVPRW.2012.6239190

M3 - Conference contribution

AN - SCOPUS:84865044286

SN - 9781467316118

T3 - IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops

SP - 15

EP - 22

BT - 2012 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2012

T2 - 2012 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2012

Y2 - 16 June 2012 through 21 June 2012

ER -

Consistency analysis and improvement for single-camera localization

Abstract

Publication series

Other

Publisher link

Find It

Other files and links

Fingerprint

Cite this