A computer-vision-based model updating strategy for shield tunnels with cracks

Xiangyu Chang; Youqi Zhang; Yuguang Fu

doi:10.1201/9781003495505-444

A computer-vision-based model updating strategy for shield tunnels with cracks

Xiangyu Chang^*, Youqi Zhang, Yuguang Fu

^*Corresponding author for this work

Nanyang Technological University

Research output: Chapter in Book/Report/Conference proceeding › Conference article in proceedings › Scientific › peer-review

110 Downloads (Pure)

Abstract

This paper presents a method that enables a finite element (FE) model of shield tunnels to update the geometric changes caused by cracks, using computer vision (CV) techniques and geometric analyses. Currently, there is a shortage of research on the automatic utilization of tunnel damage detection information for updating FE models. Thus, the objective of this study is to automate the process from tunnel damage detection to numerical analyses in order to quickly evaluate the structural capacity. The Segment Anything Model (SAM) and CV techniques are employed to determine the shapes and dimensions of both tunnel components and damages. Geometric analyses are utilized to develop algorithms that automatically transform coordinates and select elements of FE models for updating the geometric changes. A tunnel segment test is conducted to verify the effectiveness of the proposed technique and algorithms.

Original language	English
Title of host publication	Tunnelling for a Better Life - Proceedings of the ITA-AITES World Tunnel Congress, WTC 2024
Editors	Jinxiu Yan, Tarcisio Celestino, Markus Thewes, Erik Eberhardt
Publisher	CRC Press
Pages	3342-3349
Number of pages	8
ISBN (Electronic)	978-1-003-49550-5
ISBN (Print)	978-1-032-80042-4, 978-1-032-80113-1
DOIs	https://doi.org/10.1201/9781003495505-444
Publication status	Published - 2024
MoE publication type	A4 Conference publication
Event	World Tunnel Congress - Shenzhen, China Duration: 19 Apr 2024 → 25 Apr 2024

Conference

Conference	World Tunnel Congress
Abbreviated title	WTC
Country/Territory	China
City	Shenzhen
Period	19/04/2024 → 25/04/2024

Keywords

Computer vision techniques
Finite elements model
Geometric analyses
Segment anything model
Shield tunnel
Tunnel cracks

Access to Document

10.1201/9781003495505-444Licence: CC BY-NC-ND

9781040119815Final published version, 425 MBLicence: CC BY-NC-ND

Cite this

@inproceedings{29888acb69214dbfa5cefaed385765e3,

title = "A computer-vision-based model updating strategy for shield tunnels with cracks",

abstract = "This paper presents a method that enables a finite element (FE) model of shield tunnels to update the geometric changes caused by cracks, using computer vision (CV) techniques and geometric analyses. Currently, there is a shortage of research on the automatic utilization of tunnel damage detection information for updating FE models. Thus, the objective of this study is to automate the process from tunnel damage detection to numerical analyses in order to quickly evaluate the structural capacity. The Segment Anything Model (SAM) and CV techniques are employed to determine the shapes and dimensions of both tunnel components and damages. Geometric analyses are utilized to develop algorithms that automatically transform coordinates and select elements of FE models for updating the geometric changes. A tunnel segment test is conducted to verify the effectiveness of the proposed technique and algorithms.",

keywords = "Computer vision techniques, Finite elements model, Geometric analyses, Segment anything model, Shield tunnel, Tunnel cracks",

author = "Xiangyu Chang and Youqi Zhang and Yuguang Fu",

note = "Publisher Copyright: {\textcopyright} 2024 The Author(s).; World Tunnel Congress, WTC ; Conference date: 19-04-2024 Through 25-04-2024",

year = "2024",

doi = "10.1201/9781003495505-444",

language = "English",

isbn = "978-1-032-80042-4",

pages = "3342--3349",

editor = "Jinxiu Yan and Tarcisio Celestino and Markus Thewes and Erik Eberhardt",

booktitle = "Tunnelling for a Better Life - Proceedings of the ITA-AITES World Tunnel Congress, WTC 2024",

publisher = "CRC Press",

address = "United States",

}

Chang, X, Zhang, Y & Fu, Y 2024, A computer-vision-based model updating strategy for shield tunnels with cracks. in J Yan, T Celestino, M Thewes & E Eberhardt (eds), Tunnelling for a Better Life - Proceedings of the ITA-AITES World Tunnel Congress, WTC 2024. CRC Press, pp. 3342-3349, World Tunnel Congress, Shenzhen, China, 19/04/2024. https://doi.org/10.1201/9781003495505-444

A computer-vision-based model updating strategy for shield tunnels with cracks. / Chang, Xiangyu; Zhang, Youqi; Fu, Yuguang.
Tunnelling for a Better Life - Proceedings of the ITA-AITES World Tunnel Congress, WTC 2024. ed. / Jinxiu Yan; Tarcisio Celestino; Markus Thewes; Erik Eberhardt. CRC Press, 2024. p. 3342-3349.

Research output: Chapter in Book/Report/Conference proceeding › Conference article in proceedings › Scientific › peer-review

TY - GEN

T1 - A computer-vision-based model updating strategy for shield tunnels with cracks

AU - Chang, Xiangyu

AU - Zhang, Youqi

AU - Fu, Yuguang

PY - 2024

Y1 - 2024

N2 - This paper presents a method that enables a finite element (FE) model of shield tunnels to update the geometric changes caused by cracks, using computer vision (CV) techniques and geometric analyses. Currently, there is a shortage of research on the automatic utilization of tunnel damage detection information for updating FE models. Thus, the objective of this study is to automate the process from tunnel damage detection to numerical analyses in order to quickly evaluate the structural capacity. The Segment Anything Model (SAM) and CV techniques are employed to determine the shapes and dimensions of both tunnel components and damages. Geometric analyses are utilized to develop algorithms that automatically transform coordinates and select elements of FE models for updating the geometric changes. A tunnel segment test is conducted to verify the effectiveness of the proposed technique and algorithms.

AB - This paper presents a method that enables a finite element (FE) model of shield tunnels to update the geometric changes caused by cracks, using computer vision (CV) techniques and geometric analyses. Currently, there is a shortage of research on the automatic utilization of tunnel damage detection information for updating FE models. Thus, the objective of this study is to automate the process from tunnel damage detection to numerical analyses in order to quickly evaluate the structural capacity. The Segment Anything Model (SAM) and CV techniques are employed to determine the shapes and dimensions of both tunnel components and damages. Geometric analyses are utilized to develop algorithms that automatically transform coordinates and select elements of FE models for updating the geometric changes. A tunnel segment test is conducted to verify the effectiveness of the proposed technique and algorithms.

KW - Computer vision techniques

KW - Finite elements model

KW - Geometric analyses

KW - Segment anything model

KW - Shield tunnel

KW - Tunnel cracks

UR - http://www.scopus.com/inward/record.url?scp=85195466388&partnerID=8YFLogxK

UR - https://library.oapen.org/handle/20.500.12657/90204

U2 - 10.1201/9781003495505-444

DO - 10.1201/9781003495505-444

M3 - Conference article in proceedings

AN - SCOPUS:85195466388

SN - 978-1-032-80042-4

SN - 978-1-032-80113-1

SP - 3342

EP - 3349

BT - Tunnelling for a Better Life - Proceedings of the ITA-AITES World Tunnel Congress, WTC 2024

A2 - Yan, Jinxiu

A2 - Celestino, Tarcisio

A2 - Thewes, Markus

A2 - Eberhardt, Erik

PB - CRC Press

T2 - World Tunnel Congress

Y2 - 19 April 2024 through 25 April 2024

ER -

A computer-vision-based model updating strategy for shield tunnels with cracks

Abstract

Conference

Keywords

Access to Document

Other files and links

Fingerprint

SHM-TWIN / Youqi Zhang: SHM-TWIN: System for Bridge State Awareness

Cite this

A computer-vision-based model updating strategy for shield tunnels with cracks

Abstract

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Projects

SHM-TWIN / Youqi Zhang: SHM-TWIN: System for Bridge State Awareness

Cite this