A Comparison of Point Cloud Registration Techniques for on-site Disaster Data from the Surfside Structural Collapse

Ananya Ball; Robert Ladig; Pranav Goyal; John Galeotti; Howie Choset; David Merrick; Robin Murphy

doi:10.1109/SSRR56537.2022.10018779

A Comparison of Point Cloud Registration Techniques for on-site Disaster Data from the Surfside Structural Collapse

Ananya Ball
, Robert Ladig
, Pranav Goyal
, John Galeotti
, Howie Choset
, David Merrick
, Robin Murphy

Birla Institute of Technology and Science, Pilani
Florida State University
Texas A&M University
Carnegie Mellon University
Ritsumeikan University

Publikation: Kapitel i bog/rapport/konference-proceeding › Konferencebidrag i proceedings › Forskning › peer review

Abstract

3D representations of geographical surfaces in the form of dense point clouds can be a valuable tool for documenting and reconstructing a structural collapse, such as the 2021 Champlain Towers Condominium collapse in Surfside, Florida. Point cloud data reconstructed from aerial footage taken by uncrewed aerial systems at frequent intervals from a dynamic search and rescue scene poses significant challenges. Properly aligning large point clouds in this context, or registering them, poses noteworthy issues as they capture multiple regions whose geometries change over time. These regions denote dynamic features such as excavation machinery, cones marking boundaries and the structural collapse rubble itself. In this paper, the performances of commonly used point cloud registration methods for dynamic scenes present in the raw data are studied. The use of Iterative Closest Point (ICP), Rigid - Coherent Point Drift (CPD) and PointNetLK for registering dense point clouds, reconstructed sequentially over a time-frame of five days, is studied and evaluated. All methods are compared by error in performance, execution time, and robustness with a concluding analysis and a judgement of the preeminent method for the specific data at hand is provided.

Originalsprog	Engelsk
Titel	2022 - IEEE International Symposium on Safety, Security, and Rescue Robotics (SSRR)
Forlag	IEEE
Publikationsdato	2022
Sider	244-250
ISBN (Elektronisk)	9781665456807
DOI	https://doi.org/10.1109/SSRR56537.2022.10018779
Status	Udgivet - 2022
Udgivet eksternt	Ja
Begivenhed	2022 IEEE International Symposium on Safety, Security, and Rescue Robotics, SSRR 2022 - Sevilla, Spanien Varighed: 8. nov. 2022 → 10. nov. 2022

Konference

Konference	2022 IEEE International Symposium on Safety, Security, and Rescue Robotics, SSRR 2022
Land/Område	Spanien
By	Sevilla
Periode	08/11/2022 → 10/11/2022

Navn	IEEE International Symposium on Safety, Security and Rescue Robotics
ISSN	2374-3247

Bibliografisk note

Publisher Copyright:
© 2022 IEEE.

Finansiering

*Portions of this work were supported by NSF Award CMMI 2140528. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation. 1A. Bal J. Galeotti and H. Choset are with the Robotics Institute, Carnegie Mellon University, Pittsburgh, PA, USA [email protected] 2R. Ladig is with the Department of Science and Engineering, Rit-sumeikan University, Kusatsu, JAPAN 3P. Goyal is with the Birla Institute of Technology & Science - Pilani, Goa, INDIA 4D. Merrick is with the College of Social Science and Public Policy, Florida State University, FL, USA 5R. Murphy is Raytheon Professor of Computer Science and Engineering, Texas A&M University, TX, USA

Dokumenter og links

10.1109/SSRR56537.2022.10018779

SDU link

Tjek adgangen til materialet i Syddansk Universitetsbiblioteks søgemaskine

Citationsformater

@inproceedings{a207b10991b4450c935e123f34ff1b42,

title = "A Comparison of Point Cloud Registration Techniques for on-site Disaster Data from the Surfside Structural Collapse",

abstract = "3D representations of geographical surfaces in the form of dense point clouds can be a valuable tool for documenting and reconstructing a structural collapse, such as the 2021 Champlain Towers Condominium collapse in Surfside, Florida. Point cloud data reconstructed from aerial footage taken by uncrewed aerial systems at frequent intervals from a dynamic search and rescue scene poses significant challenges. Properly aligning large point clouds in this context, or registering them, poses noteworthy issues as they capture multiple regions whose geometries change over time. These regions denote dynamic features such as excavation machinery, cones marking boundaries and the structural collapse rubble itself. In this paper, the performances of commonly used point cloud registration methods for dynamic scenes present in the raw data are studied. The use of Iterative Closest Point (ICP), Rigid - Coherent Point Drift (CPD) and PointNetLK for registering dense point clouds, reconstructed sequentially over a time-frame of five days, is studied and evaluated. All methods are compared by error in performance, execution time, and robustness with a concluding analysis and a judgement of the preeminent method for the specific data at hand is provided.",

author = "Ananya Ball and Robert Ladig and Pranav Goyal and John Galeotti and Howie Choset and David Merrick and Robin Murphy",

note = "Funding Information: *Portions of this work were supported by NSF Award CMMI 2140528. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation. 1A. Bal J. Galeotti and H. Choset are with the Robotics Institute, Carnegie Mellon University, Pittsburgh, PA, USA [email protected] 2R. Ladig is with the Department of Science and Engineering, Rit-sumeikan University, Kusatsu, JAPAN 3P. Goyal is with the Birla Institute of Technology \& Science - Pilani, Goa, INDIA 4D. Merrick is with the College of Social Science and Public Policy, Florida State University, FL, USA 5R. Murphy is Raytheon Professor of Computer Science and Engineering, Texas A\&M University, TX, USA Publisher Copyright: {\textcopyright} 2022 IEEE.; 2022 IEEE International Symposium on Safety, Security, and Rescue Robotics, SSRR 2022 ; Conference date: 08-11-2022 Through 10-11-2022",

year = "2022",

doi = "10.1109/SSRR56537.2022.10018779",

language = "English",

series = "IEEE International Symposium on Safety, Security and Rescue Robotics",

publisher = "IEEE",

pages = "244--250",

booktitle = "2022 - IEEE International Symposium on Safety, Security, and Rescue Robotics (SSRR)",

address = "United States",

}

Ball, A, Ladig, R, Goyal, P, Galeotti, J, Choset, H, Merrick, D & Murphy, R 2022, A Comparison of Point Cloud Registration Techniques for on-site Disaster Data from the Surfside Structural Collapse. i 2022 - IEEE International Symposium on Safety, Security, and Rescue Robotics (SSRR). IEEE, IEEE International Symposium on Safety, Security and Rescue Robotics, s. 244-250, 2022 IEEE International Symposium on Safety, Security, and Rescue Robotics, SSRR 2022, Sevilla, Spanien, 08/11/2022. https://doi.org/10.1109/SSRR56537.2022.10018779

A Comparison of Point Cloud Registration Techniques for on-site Disaster Data from the Surfside Structural Collapse. / Ball, Ananya; Ladig, Robert; Goyal, Pranav et al.
2022 - IEEE International Symposium on Safety, Security, and Rescue Robotics (SSRR). IEEE, 2022. s. 244-250 (IEEE International Symposium on Safety, Security and Rescue Robotics).

Publikation: Kapitel i bog/rapport/konference-proceeding › Konferencebidrag i proceedings › Forskning › peer review

TY - GEN

T1 - A Comparison of Point Cloud Registration Techniques for on-site Disaster Data from the Surfside Structural Collapse

AU - Ball, Ananya

AU - Ladig, Robert

AU - Goyal, Pranav

AU - Galeotti, John

AU - Choset, Howie

AU - Merrick, David

AU - Murphy, Robin

N1 - Funding Information: *Portions of this work were supported by NSF Award CMMI 2140528. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation. 1A. Bal J. Galeotti and H. Choset are with the Robotics Institute, Carnegie Mellon University, Pittsburgh, PA, USA [email protected] 2R. Ladig is with the Department of Science and Engineering, Rit-sumeikan University, Kusatsu, JAPAN 3P. Goyal is with the Birla Institute of Technology & Science - Pilani, Goa, INDIA 4D. Merrick is with the College of Social Science and Public Policy, Florida State University, FL, USA 5R. Murphy is Raytheon Professor of Computer Science and Engineering, Texas A&M University, TX, USA Publisher Copyright: © 2022 IEEE.

PY - 2022

Y1 - 2022

N2 - 3D representations of geographical surfaces in the form of dense point clouds can be a valuable tool for documenting and reconstructing a structural collapse, such as the 2021 Champlain Towers Condominium collapse in Surfside, Florida. Point cloud data reconstructed from aerial footage taken by uncrewed aerial systems at frequent intervals from a dynamic search and rescue scene poses significant challenges. Properly aligning large point clouds in this context, or registering them, poses noteworthy issues as they capture multiple regions whose geometries change over time. These regions denote dynamic features such as excavation machinery, cones marking boundaries and the structural collapse rubble itself. In this paper, the performances of commonly used point cloud registration methods for dynamic scenes present in the raw data are studied. The use of Iterative Closest Point (ICP), Rigid - Coherent Point Drift (CPD) and PointNetLK for registering dense point clouds, reconstructed sequentially over a time-frame of five days, is studied and evaluated. All methods are compared by error in performance, execution time, and robustness with a concluding analysis and a judgement of the preeminent method for the specific data at hand is provided.

AB - 3D representations of geographical surfaces in the form of dense point clouds can be a valuable tool for documenting and reconstructing a structural collapse, such as the 2021 Champlain Towers Condominium collapse in Surfside, Florida. Point cloud data reconstructed from aerial footage taken by uncrewed aerial systems at frequent intervals from a dynamic search and rescue scene poses significant challenges. Properly aligning large point clouds in this context, or registering them, poses noteworthy issues as they capture multiple regions whose geometries change over time. These regions denote dynamic features such as excavation machinery, cones marking boundaries and the structural collapse rubble itself. In this paper, the performances of commonly used point cloud registration methods for dynamic scenes present in the raw data are studied. The use of Iterative Closest Point (ICP), Rigid - Coherent Point Drift (CPD) and PointNetLK for registering dense point clouds, reconstructed sequentially over a time-frame of five days, is studied and evaluated. All methods are compared by error in performance, execution time, and robustness with a concluding analysis and a judgement of the preeminent method for the specific data at hand is provided.

U2 - 10.1109/SSRR56537.2022.10018779

DO - 10.1109/SSRR56537.2022.10018779

M3 - Article in proceedings

AN - SCOPUS:85147550091

T3 - IEEE International Symposium on Safety, Security and Rescue Robotics

SP - 244

EP - 250

BT - 2022 - IEEE International Symposium on Safety, Security, and Rescue Robotics (SSRR)

PB - IEEE

T2 - 2022 IEEE International Symposium on Safety, Security, and Rescue Robotics, SSRR 2022

Y2 - 8 November 2022 through 10 November 2022

ER -

Ball A, Ladig R, Goyal P, Galeotti J, Choset H, Merrick D et al. A Comparison of Point Cloud Registration Techniques for on-site Disaster Data from the Surfside Structural Collapse. I 2022 - IEEE International Symposium on Safety, Security, and Rescue Robotics (SSRR). IEEE. 2022. s. 244-250. (IEEE International Symposium on Safety, Security and Rescue Robotics). doi: 10.1109/SSRR56537.2022.10018779

A Comparison of Point Cloud Registration Techniques for on-site Disaster Data from the Surfside Structural Collapse

Abstract

Konference

Bibliografisk note

Finansiering

Dokumenter og links

SDU link

Fingeraftryk

Citationsformater