R3P: real-time RGB-D registration pipeline

Hani Javan Hemmat; Egor Bondarev; Peter H.N. de With

doi:10.1007/978-3-319-48680-2_34

R³P: real-time RGB-D registration pipeline

Hani Javan Hemmat, Egor Bondarev, Peter H.N. de With

Video Coding & Architectures

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

Abstract

Applications based on colored 3-D data sequences suffer from lack of efficient algorithms for transformation estimation and key points extraction to perform accurate registration and sensor localization either in the 2-D or 3-D domain. Therefore, we propose a real-time RGB-D registration pipeline, named R³P, presented in processing layers. In this paper, we present an evaluation of several algorithm combinations for each layer, to optimize the registration and sensor localization for specific applications. The resulting dynamic reconfigurability of R³P makes it suitable as a front-end system for any SLAM reconstruction algorithm. Evaluation results on several public datasets reveal that R³P delivers real-time registration with 59 fps and high accuracy with the relative pose error (for a time span of 40 frames) for rotation and translation of 0.5ï¿½ and 8 mm, respectively. All the heterogeneous dataset and implementations are publicly available under an open-source license [21].

Original language	English
Title of host publication	Advanced Concepts for Intelligent Vision Systems
Subtitle of host publication	17th International Conference, ACIVS 2016, Lecce, Italy, October 24-27, 2016, Proceedings
Editors	J. Blanc-Talon, C. Distante, W. Philips, D. Popescu, P. Scheunders
Place of Publication	Dordrecht
Publisher	Springer
Pages	385-397
Number of pages	13
ISBN (Electronic)	978-3-319-48680-2
ISBN (Print)	978-3-319-48679-6
DOIs	https://doi.org/10.1007/978-3-319-48680-2_34
Publication status	Published - 1 Jan 2016
Event	17th International Conference on Advanced Concepts for Intelligent Vision Systems (ACIVS 2016) - Lecce, Italy Duration: 24 Oct 2016 → 27 Oct 2016 Conference number: 17

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	10016 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	17th International Conference on Advanced Concepts for Intelligent Vision Systems (ACIVS 2016)
Abbreviated title	ACIVS 2016
Country/Territory	Italy
City	Lecce
Period	24/10/16 → 27/10/16

Keywords

Heterogeneous 3-D reconstruction
Real-time
RGB-D registration
Simultaneously localization and mapping (SLAM)

Access to Document

10.1007/978-3-319-48680-2_34

Cite this

Hemmat, H. J., Bondarev, E., & de With, P. H. N. (2016). R³P: real-time RGB-D registration pipeline. In J. Blanc-Talon, C. Distante, W. Philips, D. Popescu, & P. Scheunders (Eds.), Advanced Concepts for Intelligent Vision Systems: 17th International Conference, ACIVS 2016, Lecce, Italy, October 24-27, 2016, Proceedings (pp. 385-397). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10016 LNCS). Springer. https://doi.org/10.1007/978-3-319-48680-2_34

Hemmat, Hani Javan ; Bondarev, Egor ; de With, Peter H.N. / R³P : real-time RGB-D registration pipeline. Advanced Concepts for Intelligent Vision Systems: 17th International Conference, ACIVS 2016, Lecce, Italy, October 24-27, 2016, Proceedings. editor / J. Blanc-Talon ; C. Distante ; W. Philips ; D. Popescu ; P. Scheunders. Dordrecht : Springer, 2016. pp. 385-397 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{585f20e871b14187b7352e68ead7645c,

title = "R3P: real-time RGB-D registration pipeline",

abstract = "Applications based on colored 3-D data sequences suffer from lack of efficient algorithms for transformation estimation and key points extraction to perform accurate registration and sensor localization either in the 2-D or 3-D domain. Therefore, we propose a real-time RGB-D registration pipeline, named R3P, presented in processing layers. In this paper, we present an evaluation of several algorithm combinations for each layer, to optimize the registration and sensor localization for specific applications. The resulting dynamic reconfigurability of R3P makes it suitable as a front-end system for any SLAM reconstruction algorithm. Evaluation results on several public datasets reveal that R3P delivers real-time registration with 59 fps and high accuracy with the relative pose error (for a time span of 40 frames) for rotation and translation of 0.5{\"i}¿½ and 8 mm, respectively. All the heterogeneous dataset and implementations are publicly available under an open-source license [21].",

keywords = "Heterogeneous 3-D reconstruction, Real-time, RGB-D registration, Simultaneously localization and mapping (SLAM)",

author = "Hemmat, {Hani Javan} and Egor Bondarev and {de With}, {Peter H.N.}",

year = "2016",

month = jan,

day = "1",

doi = "10.1007/978-3-319-48680-2_34",

language = "English",

isbn = "978-3-319-48679-6",

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

publisher = "Springer",

pages = "385--397",

editor = "J. Blanc-Talon and C. Distante and W. Philips and D. Popescu and P. Scheunders",

booktitle = "Advanced Concepts for Intelligent Vision Systems",

address = "Germany",

note = "17th International Conference on Advanced Concepts for Intelligent Vision Systems (ACIVS 2016), ACIVS 2016 ; Conference date: 24-10-2016 Through 27-10-2016",

}

Hemmat, HJ, Bondarev, E & de With, PHN 2016, R³P: real-time RGB-D registration pipeline. in J Blanc-Talon, C Distante, W Philips, D Popescu & P Scheunders (eds), Advanced Concepts for Intelligent Vision Systems: 17th International Conference, ACIVS 2016, Lecce, Italy, October 24-27, 2016, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 10016 LNCS, Springer, Dordrecht, pp. 385-397, 17th International Conference on Advanced Concepts for Intelligent Vision Systems (ACIVS 2016), Lecce, Italy, 24/10/16. https://doi.org/10.1007/978-3-319-48680-2_34

R³P: real-time RGB-D registration pipeline. / Hemmat, Hani Javan; Bondarev, Egor ; de With, Peter H.N.
Advanced Concepts for Intelligent Vision Systems: 17th International Conference, ACIVS 2016, Lecce, Italy, October 24-27, 2016, Proceedings. ed. / J. Blanc-Talon; C. Distante; W. Philips; D. Popescu; P. Scheunders. Dordrecht: Springer, 2016. p. 385-397 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10016 LNCS).

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

TY - GEN

T1 - R3P

T2 - 17th International Conference on Advanced Concepts for Intelligent Vision Systems (ACIVS 2016)

AU - Hemmat, Hani Javan

AU - Bondarev, Egor

AU - de With, Peter H.N.

N1 - Conference code: 17

PY - 2016/1/1

Y1 - 2016/1/1

N2 - Applications based on colored 3-D data sequences suffer from lack of efficient algorithms for transformation estimation and key points extraction to perform accurate registration and sensor localization either in the 2-D or 3-D domain. Therefore, we propose a real-time RGB-D registration pipeline, named R3P, presented in processing layers. In this paper, we present an evaluation of several algorithm combinations for each layer, to optimize the registration and sensor localization for specific applications. The resulting dynamic reconfigurability of R3P makes it suitable as a front-end system for any SLAM reconstruction algorithm. Evaluation results on several public datasets reveal that R3P delivers real-time registration with 59 fps and high accuracy with the relative pose error (for a time span of 40 frames) for rotation and translation of 0.5ï¿½ and 8 mm, respectively. All the heterogeneous dataset and implementations are publicly available under an open-source license [21].

AB - Applications based on colored 3-D data sequences suffer from lack of efficient algorithms for transformation estimation and key points extraction to perform accurate registration and sensor localization either in the 2-D or 3-D domain. Therefore, we propose a real-time RGB-D registration pipeline, named R3P, presented in processing layers. In this paper, we present an evaluation of several algorithm combinations for each layer, to optimize the registration and sensor localization for specific applications. The resulting dynamic reconfigurability of R3P makes it suitable as a front-end system for any SLAM reconstruction algorithm. Evaluation results on several public datasets reveal that R3P delivers real-time registration with 59 fps and high accuracy with the relative pose error (for a time span of 40 frames) for rotation and translation of 0.5ï¿½ and 8 mm, respectively. All the heterogeneous dataset and implementations are publicly available under an open-source license [21].

KW - Heterogeneous 3-D reconstruction

KW - Real-time

KW - RGB-D registration

KW - Simultaneously localization and mapping (SLAM)

UR - http://www.scopus.com/inward/record.url?scp=84994410884&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-48680-2_34

DO - 10.1007/978-3-319-48680-2_34

M3 - Conference contribution

AN - SCOPUS:84994410884

SN - 978-3-319-48679-6

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

SP - 385

EP - 397

BT - Advanced Concepts for Intelligent Vision Systems

A2 - Blanc-Talon, J.

A2 - Distante, C.

A2 - Philips, W.

A2 - Popescu, D.

A2 - Scheunders, P.

PB - Springer

CY - Dordrecht

Y2 - 24 October 2016 through 27 October 2016

ER -

Hemmat HJ, Bondarev E , de With PHN. R³P: real-time RGB-D registration pipeline. In Blanc-Talon J, Distante C, Philips W, Popescu D, Scheunders P, editors, Advanced Concepts for Intelligent Vision Systems: 17th International Conference, ACIVS 2016, Lecce, Italy, October 24-27, 2016, Proceedings. Dordrecht: Springer. 2016. p. 385-397. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-319-48680-2_34