An automatic defect classification and segmentation method on three-dimensional point clouds for sewer pipes

•An 3D inspection method on point clouds for pipe defects was proposed.•The network structure was optimized to improve the inspection accuracy.•The training strategies were improved to stabilize training and avoid overfitting.•Two data augmentation methods are used to facilitate training. With the d...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:Tunnelling and underground space technology Ročník 143; s. 105480
Hlavní autoři: Wang, Niannian, Ma, Duo, Du, Xueming, Li, Bin, Di, Danyang, Pang, Gaozhao, Duan, Yihang
Médium: Journal Article
Jazyk:angličtina
Vydáno: Elsevier Ltd 01.01.2024
Témata:
ISSN:0886-7798, 1878-4364
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Abstract •An 3D inspection method on point clouds for pipe defects was proposed.•The network structure was optimized to improve the inspection accuracy.•The training strategies were improved to stabilize training and avoid overfitting.•Two data augmentation methods are used to facilitate training. With the development of deep learning (DL), sewer pipe inspection on two-dimensional (2D) images has achieved remarkable accuracy. However, extracting defect measurements from these 2D images is challenging due to the curved nature of pipes and the lack of depth information. Point clouds can restore the three-dimensional (3D) information of objects. To effectively identify defects in disordered and sparse point clouds, a 3D sewer pipe classification and segmentation method was proposed. In the encoder, the original point clouds are sampled and grouped and the local features in the clusters are extracted by two symmetric functions (1 × 1 convolution and the maximization function) to process the points with permutation invariance. In the decoder, the multi-scaling abstract features are upsampled using feature pyramid network (FPN) to predict the category of each point. Especially, the network structure and training strategy of the inspection method is optimized to improve the inspection accuracy. Furthermore, two data augmentation methods, namely random scaling and point jitter, are used to increase the data volume. An ablation experiment shows that the optimization of network structure can effectively improve the performance of the inspection model and the novel training strategies can stabilize the training process and prevent overfitting. Comparison among the state-of-the-art networks demonstrates that the proposed segmentation model attains the highest mIoU of 94.15 %, which is improved by 11.46 % with the optimization of network structure and training strategy. For the classification task, the F1 score and accuracy of the established model are 6.79 % and 5.46 % higher than PointNet++, respectively. These results signify the high-accuracy defect inspection capability of our proposed method on 3D point clouds of sewer pipelines.
AbstractList •An 3D inspection method on point clouds for pipe defects was proposed.•The network structure was optimized to improve the inspection accuracy.•The training strategies were improved to stabilize training and avoid overfitting.•Two data augmentation methods are used to facilitate training. With the development of deep learning (DL), sewer pipe inspection on two-dimensional (2D) images has achieved remarkable accuracy. However, extracting defect measurements from these 2D images is challenging due to the curved nature of pipes and the lack of depth information. Point clouds can restore the three-dimensional (3D) information of objects. To effectively identify defects in disordered and sparse point clouds, a 3D sewer pipe classification and segmentation method was proposed. In the encoder, the original point clouds are sampled and grouped and the local features in the clusters are extracted by two symmetric functions (1 × 1 convolution and the maximization function) to process the points with permutation invariance. In the decoder, the multi-scaling abstract features are upsampled using feature pyramid network (FPN) to predict the category of each point. Especially, the network structure and training strategy of the inspection method is optimized to improve the inspection accuracy. Furthermore, two data augmentation methods, namely random scaling and point jitter, are used to increase the data volume. An ablation experiment shows that the optimization of network structure can effectively improve the performance of the inspection model and the novel training strategies can stabilize the training process and prevent overfitting. Comparison among the state-of-the-art networks demonstrates that the proposed segmentation model attains the highest mIoU of 94.15 %, which is improved by 11.46 % with the optimization of network structure and training strategy. For the classification task, the F1 score and accuracy of the established model are 6.79 % and 5.46 % higher than PointNet++, respectively. These results signify the high-accuracy defect inspection capability of our proposed method on 3D point clouds of sewer pipelines.
ArticleNumber 105480
Author Duan, Yihang
Di, Danyang
Pang, Gaozhao
Li, Bin
Ma, Duo
Wang, Niannian
Du, Xueming
Author_xml – sequence: 1
  givenname: Niannian
  surname: Wang
  fullname: Wang, Niannian
– sequence: 2
  givenname: Duo
  surname: Ma
  fullname: Ma, Duo
  email: md13255961153@gs.zzu.edu.cn
– sequence: 3
  givenname: Xueming
  surname: Du
  fullname: Du, Xueming
– sequence: 4
  givenname: Bin
  surname: Li
  fullname: Li, Bin
– sequence: 5
  givenname: Danyang
  surname: Di
  fullname: Di, Danyang
– sequence: 6
  givenname: Gaozhao
  surname: Pang
  fullname: Pang, Gaozhao
– sequence: 7
  givenname: Yihang
  surname: Duan
  fullname: Duan, Yihang
BookMark eNp9kM1qxCAURqVModNpX6ArXyBTY4wx0M0w9A8GumnX4ui145DEoE5L376GdNXFrJTv41zuPddoMfgBELorybokJb8_rtMppjUltMpBzQS5QMtSNKJgFWcLtCRC8KJpWnGFrmM8EkJqStsl6jYDVqfke5WcxgYs6IR1p2J01ukc-twPBkf47GFIc9BDOniD8y8dAkBhXO5iblSHR--GaYI_mYitD5n8hoBHN0K8QZdWdRFu_94V-nh6fN--FLu359ftZlfoipBUNLylivG9EKpinGiVt-WW16xqjW0Vp9Aw3nDD1H5fl5QSa5k1xLaWtNCUrFohMc_VwccYwErt5tVTUK6TJZGTNXmUkzU5WZOztYzSf-gYXK_Cz3noYYYgH_XlIMioHQwajAvZpzTencN_AeuSiuY
CitedBy_id crossref_primary_10_1016_j_measurement_2024_115954
crossref_primary_10_1111_mice_13241
crossref_primary_10_1061_JCCEE5_CPENG_6549
crossref_primary_10_1016_j_inffus_2024_102575
crossref_primary_10_1155_2024_9217395
crossref_primary_10_1016_j_tust_2025_106755
crossref_primary_10_1016_j_measurement_2025_117434
crossref_primary_10_3390_rs17010133
crossref_primary_10_1016_j_tust_2024_106130
crossref_primary_10_1016_j_autcon_2024_105769
crossref_primary_10_1016_j_aei_2024_102936
crossref_primary_10_1016_j_autcon_2025_106098
crossref_primary_10_1016_j_undsp_2025_01_007
crossref_primary_10_3390_s24237557
crossref_primary_10_1016_j_tust_2024_106195
crossref_primary_10_3390_s24237786
crossref_primary_10_1016_j_measurement_2025_117370
crossref_primary_10_3390_photonics11070635
crossref_primary_10_1061_JCCEE5_CPENG_6672
crossref_primary_10_3390_w17182781
crossref_primary_10_1016_j_asoc_2024_111622
crossref_primary_10_1109_TIM_2025_3575961
crossref_primary_10_3390_app15073518
Cites_doi 10.1016/j.autcon.2021.103874
10.1111/mice.12970
10.1016/j.autcon.2022.104367
10.1016/j.tust.2022.104861
10.1109/TITS.2021.3054026
10.1016/j.autcon.2022.104555
10.1109/JSTARS.2018.2817227
10.1061/(ASCE)1076-0342(2005)11:3(165)
10.1016/j.autcon.2014.12.015
10.1016/j.autcon.2022.104494
10.1016/j.autcon.2021.103755
10.3390/buildings12020213
10.1016/j.autcon.2022.104285
10.1016/j.tust.2022.104403
10.1016/j.tust.2021.103840
10.1109/CVPR.2016.90
10.1016/j.aei.2020.101200
10.1016/j.autcon.2022.104519
10.1016/j.autcon.2017.11.004
10.1016/j.undsp.2021.08.004
10.1016/j.enggeo.2018.03.020
10.1109/MGRS.2019.2937630
10.1016/j.autcon.2022.104167
10.5220/0010207908910900
10.1016/j.tust.2022.104965
10.1016/j.asoc.2018.11.016
10.1016/j.autcon.2020.103383
10.1016/S0926-5805(97)00071-X
10.1111/mice.12918
10.3390/w11102101
10.1016/j.autcon.2020.103289
10.1007/s10044-013-0355-5
10.1016/j.autcon.2021.103992
10.1016/j.tust.2022.104761
10.1016/j.autcon.2018.03.028
10.1016/j.aei.2018.05.005
10.1016/j.autcon.2022.104163
10.1016/j.conbuildmat.2021.125385
ContentType Journal Article
Copyright 2023 Elsevier Ltd
Copyright_xml – notice: 2023 Elsevier Ltd
DBID AAYXX
CITATION
DOI 10.1016/j.tust.2023.105480
DatabaseName CrossRef
DatabaseTitle CrossRef
DatabaseTitleList
DeliveryMethod fulltext_linktorsrc
Discipline Engineering
EISSN 1878-4364
ExternalDocumentID 10_1016_j_tust_2023_105480
S088677982300500X
GroupedDBID --K
--M
.~1
0R~
123
1B1
1RT
1~.
1~5
29Q
4.4
457
4G.
5VS
7-5
71M
8P~
9JN
AACTN
AAEDT
AAEDW
AAIAV
AAIKJ
AAKOC
AALRI
AAOAW
AAQFI
AAQXK
AAXUO
ABFNM
ABJNI
ABMAC
ABQEM
ABQYD
ABXDB
ABYKQ
ACDAQ
ACGFS
ACIWK
ACLVX
ACNNM
ACRLP
ACSBN
ADBBV
ADEZE
ADMUD
ADTZH
AEBSH
AECPX
AEKER
AENEX
AFKWA
AFTJW
AGHFR
AGUBO
AGYEJ
AHHHB
AHJVU
AIEXJ
AIKHN
AITUG
AJBFU
AJOXV
ALMA_UNASSIGNED_HOLDINGS
AMFUW
AMRAJ
ASPBG
ATOGT
AVWKF
AXJTR
AZFZN
BJAXD
BKOJK
BLXMC
CS3
DU5
EBS
EFJIC
EFLBG
EJD
EO8
EO9
EP2
EP3
FDB
FEDTE
FGOYB
FIRID
FNPLU
FYGXN
G-2
G-Q
GBLVA
HMA
HVGLF
HZ~
IHE
IMUCA
J1W
JJJVA
KOM
LY3
LY7
M41
MO0
N9A
O-L
O9-
OAUVE
OZT
P-8
P-9
P2P
PC.
Q38
R2-
RIG
ROL
RPZ
SDF
SDG
SEP
SES
SET
SEW
SPC
SPCBC
SSE
SST
SSZ
T5K
WUQ
ZMT
~02
~G-
9DU
AATTM
AAXKI
AAYWO
AAYXX
ABWVN
ACLOT
ACRPL
ACVFH
ADCNI
ADNMO
AEIPS
AEUPX
AFJKZ
AFPUW
AGQPQ
AIGII
AIIUN
AKBMS
AKRWK
AKYEP
ANKPU
APXCP
CITATION
EFKBS
~HD
ID FETCH-LOGICAL-c300t-7692a46b88a3460ca0006f65439df9a62e74676d4abb51220ff4fd0f9f09e7143
ISICitedReferencesCount 27
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=001100271500001&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN 0886-7798
IngestDate Tue Nov 18 22:35:44 EST 2025
Sat Nov 29 07:10:21 EST 2025
Fri Feb 23 02:35:54 EST 2024
IsPeerReviewed true
IsScholarly true
Keywords Deep learning
3D segmentation
Point cloud
Sewer pipeline
Defect inspection
Language English
LinkModel OpenURL
MergedId FETCHMERGED-LOGICAL-c300t-7692a46b88a3460ca0006f65439df9a62e74676d4abb51220ff4fd0f9f09e7143
ParticipantIDs crossref_citationtrail_10_1016_j_tust_2023_105480
crossref_primary_10_1016_j_tust_2023_105480
elsevier_sciencedirect_doi_10_1016_j_tust_2023_105480
PublicationCentury 2000
PublicationDate January 2024
2024-01-00
PublicationDateYYYYMMDD 2024-01-01
PublicationDate_xml – month: 01
  year: 2024
  text: January 2024
PublicationDecade 2020
PublicationTitle Tunnelling and underground space technology
PublicationYear 2024
Publisher Elsevier Ltd
Publisher_xml – name: Elsevier Ltd
References Wirahadikusumah, Abraham, Iseley, Prasanth (b0145) 1998; 7
Meng, S., Gao, Z., Zhou, Y., He, B., Djerrad, A., 2022. Real‐time automatic crack detection method based on drone. Comput.-Aided Civil Infrastruct. Eng.
Zhang, Du, Tannant, Zhu, Zheng (b0200) 2018; 239
Zhang, Hu, Ai (b0205) 2019; 11
Zhao, S., Kang, F., Li, J., 2022. Concrete dam damage detection and localisation based on YOLOv5s-HSC and photogrammetric 3D reconstruction. Autom. Constr. 143.
Liu, H., Yue, Y., Liu, C., Spencer, B.F., Cui, J., 2022. Automatic recognition and localization of underground pipelines in GPR B-scans using a deep learning model. Tunn. Undergr. Space Technol.
Xia, T., Yang, J., Chen, L., 2022. Automated semantic segmentation of bridge point cloud based on local descriptor and machine learning. Autom. Constr. 133.
Yin, C., Wang, B., Gan, V.J.L., Wang, M., Cheng, J.C.P., 2021. Automated semantic segmentation of industrial point clouds using ResPointNet++. Autom. Constr. 130.
Zhai, Moore (b0195) 2023; 133
Ma, Liu, Fang, Wang, Zhang, Li, Dong (b0075) 2021; 312
Xu, Yao, Tuttas, Hoegner, Stilla (b0170) 2018; 11
A.C. Wilson R. Roelofs M. Stern N. Srebro B. Recht The Marginal Value of Adaptive Gradient Methods in Machine Learning 2017.
He, K., Zhang, X., Ren, S., Sun, J., 2016. Deep residual learning for image recognition. In: Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, pp. 770-778.
Ma, J.W., Leite, F., 2022. Performance boosting of conventional deep learning-based semantic segmentation leveraging unsupervised clustering. Autom. Constr. 136.
Moeslund, T., Nikolov, I., Henriksen, K., Lynge, M., Allahham, M., Haurum, J., 2021. Sewer defect classification using synthetic point clouds. In: Proceedings of the 16th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications, pp. 891-900.
Wu, Liu, He (b0150) 2013; 18
Wang, Cheng (b0130) 2020; 35
Koo, Jung, Yu, Kim (b0040) 2021; 8
Shehab, Moselhi (b0115) 2005; 11
Zhou, Y., Ji, A., Zhang, L., 2022b. Sewer defect detection from 3D point clouds using a transformer-based deep learning model. Autom. Constr. 136.
Reyes-Acosta, Lopez-Juarez, Osorio-Comparan, Lefranc (b0110) 2019; 75
Xie, Tian, Zhu (b0160) 2020; 8
Ahmed, Ashfaque, Ulhaq, Mathavan, Kamal, Rahman (b0005) 2022; 23
Koo, Jung, Yu (b0035) 2021; 47
Pan, Zheng, Guo, Lv (b0100) 2020; 119
Oh, Lee, Kim, Kim, Cho (b0095) 2021; 13
Insa-Iglesias, M., Jenkins, M.D., Morison, G., 2021. 3D visual inspection system framework for structural condition monitoring and analysis. Autom. Constr. 128.
Sun, J., Peng, B., Wang, C.C., Chen, K., Zhong, B., Wu, J., 2022. Building displacement measurement and analysis based on UAV images. Autom. Constr. 140.
Fang, X., Li, Q., Zhu, J., Chen, Z., Zhang, D., Wu, K., Ding, K., Li, Q., 2022. Sewer defect instance segmentation, localization, and 3D reconstruction for sewer floating capsule robots. Autom. Constr. 142.
Huang, Wang, Fang, Liu, Pang (b0025) 2022; 12
Ma, D., Fang, H., Wang, N., Lu, H., Matthews, J., Zhang, C., 2023. Transformer‐optimized generation, detection, and tracking network for images with drainage pipeline defects. Comput.-Aided Civil Infrastruct. Eng.
United States Department of Transportation, Pipeline Incident 20 Year Trends, https://www.phmsa.dot.gov/data-and-statistics/pipeline/pipeline-incident-20-year-trends.
Xu, Liu (b0165) 2021; 37
Pang, Wang, Fang, Liu, Huang (b0105) 2022; 12
Xue, Shi, Jia, Huang (b0175) 2022; 7
Li, Fang, Yang, Zhang, Du, Wang, Guo (b0050) 2022; 130
Yoon, Wang, Sohn (b0190) 2018; 86
Dimitrov, Golparvar-Fard (b0010) 2015; 51
Yang, X., del Rey Castillo, E., Zou, Y., Wotherspoon, L., Tan, Y., 2022. Automated semantic segmentation of bridge components from large-scale point clouds using a weighted superpoint graph. Autom. Constr. 142.
Kumar, Abraham, Jahanshahi, Iseley, Starr (b0045) 2018; 91
Zhou, Situ, Teng, Liu, Chen, Chen (b0225) 2022; 123
Zhang, Zhang, Fu, Ozevin, Yuan (b0210) 2021
Wang, Luo, Cheng (b0135) 2021; 110
Ma, Liu (b0070) 2018; 37
Meijer, D., Luimes, R., Knobbe, A., Bäck, T., 2022. Anomaly detection in urban drainage with stereovision. Autom. Constr. 139.
Zuo, X., Dai, B., Shan, Y., Shen, J., Hu, C., Huang, S., 2020. Classifying cracks at sub-class level in closed circuit television sewer inspection videos. Autom. Constr. 118.
Yoon (10.1016/j.tust.2023.105480_b0190) 2018; 86
Pan (10.1016/j.tust.2023.105480_b0100) 2020; 119
10.1016/j.tust.2023.105480_b0215
10.1016/j.tust.2023.105480_b0015
Oh (10.1016/j.tust.2023.105480_b0095) 2021; 13
10.1016/j.tust.2023.105480_b0020
Li (10.1016/j.tust.2023.105480_b0050) 2022; 130
Xu (10.1016/j.tust.2023.105480_b0170) 2018; 11
10.1016/j.tust.2023.105480_b0185
10.1016/j.tust.2023.105480_b0065
Pang (10.1016/j.tust.2023.105480_b0105) 2022; 12
10.1016/j.tust.2023.105480_b0120
10.1016/j.tust.2023.105480_b0085
10.1016/j.tust.2023.105480_b0140
Wang (10.1016/j.tust.2023.105480_b0130) 2020; 35
Wang (10.1016/j.tust.2023.105480_b0135) 2021; 110
Zhang (10.1016/j.tust.2023.105480_b0205) 2019; 11
10.1016/j.tust.2023.105480_b0220
Koo (10.1016/j.tust.2023.105480_b0040) 2021; 8
Kumar (10.1016/j.tust.2023.105480_b0045) 2018; 91
Shehab (10.1016/j.tust.2023.105480_b0115) 2005; 11
10.1016/j.tust.2023.105480_b0060
10.1016/j.tust.2023.105480_b0080
Reyes-Acosta (10.1016/j.tust.2023.105480_b0110) 2019; 75
10.1016/j.tust.2023.105480_b0180
Ma (10.1016/j.tust.2023.105480_b0070) 2018; 37
Koo (10.1016/j.tust.2023.105480_b0035) 2021; 47
Xue (10.1016/j.tust.2023.105480_b0175) 2022; 7
Ahmed (10.1016/j.tust.2023.105480_b0005) 2022; 23
Zhai (10.1016/j.tust.2023.105480_b0195) 2023; 133
Zhou (10.1016/j.tust.2023.105480_b0225) 2022; 123
Ma (10.1016/j.tust.2023.105480_b0075) 2021; 312
10.1016/j.tust.2023.105480_b0125
Zhang (10.1016/j.tust.2023.105480_b0210) 2021
Wu (10.1016/j.tust.2023.105480_b0150) 2013; 18
Zhang (10.1016/j.tust.2023.105480_b0200) 2018; 239
Dimitrov (10.1016/j.tust.2023.105480_b0010) 2015; 51
10.1016/j.tust.2023.105480_b0230
10.1016/j.tust.2023.105480_b0030
Huang (10.1016/j.tust.2023.105480_b0025) 2022; 12
Xie (10.1016/j.tust.2023.105480_b0160) 2020; 8
10.1016/j.tust.2023.105480_b0055
Wirahadikusumah (10.1016/j.tust.2023.105480_b0145) 1998; 7
10.1016/j.tust.2023.105480_b0155
10.1016/j.tust.2023.105480_b0090
Xu (10.1016/j.tust.2023.105480_b0165) 2021; 37
References_xml – volume: 8
  start-page: 38
  year: 2020
  end-page: 59
  ident: b0160
  article-title: Linking Points With Labels in 3D: A Review of Point Cloud Semantic Segmentation
  publication-title: IEEE Geosci. Remote Sens. Mag.
– volume: 130
  year: 2022
  ident: b0050
  article-title: Impact of erosion voids and internal corrosion on concrete pipes under traffic loads
  publication-title: Tunn. Undergr. Space Technol.
– volume: 23
  start-page: 4685
  year: 2022
  end-page: 4694
  ident: b0005
  article-title: Pothole 3D reconstruction with a novel imaging system and structure from motion techniques
  publication-title: IEEE Trans. Intell. Transp. Syst.
– volume: 133
  year: 2023
  ident: b0195
  article-title: Axial stresses in pressure pipe liners spanning joints with initial gap, opening as a result of differential ground movements
  publication-title: Tunn. Undergr. Space Technol.
– reference: Fang, X., Li, Q., Zhu, J., Chen, Z., Zhang, D., Wu, K., Ding, K., Li, Q., 2022. Sewer defect instance segmentation, localization, and 3D reconstruction for sewer floating capsule robots. Autom. Constr. 142.
– volume: 35
  start-page: 162
  year: 2020
  end-page: 177
  ident: b0130
  article-title: A unified convolutional neural network integrated with conditional random field for pipe defect segmentation. Comput.-Aided Civil Infrastruct
  publication-title: Eng.
– reference: Zhou, Y., Ji, A., Zhang, L., 2022b. Sewer defect detection from 3D point clouds using a transformer-based deep learning model. Autom. Constr. 136.
– volume: 123
  start-page: 1
  year: 2022
  end-page: 14
  ident: b0225
  article-title: Automatic sewer defect detection and severity quantification based on pixel-level semantic segmentation
  publication-title: Tunn. Undergr. Space Technol.
– volume: 51
  start-page: 32
  year: 2015
  end-page: 45
  ident: b0010
  article-title: Segmentation of building point cloud models including detailed architectural/structural features and MEP systems
  publication-title: Autom. Constr.
– volume: 11
  start-page: 2101
  year: 2019
  ident: b0205
  article-title: A 3D reconstruction pipeline of urban drainage pipes based on multiviewImage matching using low-cost panoramic video cameras
  publication-title: Water
– volume: 110
  start-page: 1
  year: 2021
  end-page: 20
  ident: b0135
  article-title: Towards an automated condition assessment framework of underground sewer pipes based on closed-circuit television (CCTV) images
  publication-title: Tunn. Undergr. Space Technol.
– reference: Liu, H., Yue, Y., Liu, C., Spencer, B.F., Cui, J., 2022. Automatic recognition and localization of underground pipelines in GPR B-scans using a deep learning model. Tunn. Undergr. Space Technol.
– volume: 7
  start-page: 259
  year: 1998
  end-page: 270
  ident: b0145
  article-title: Assessment technologies for sewer system rehabilitation
  publication-title: Autom. Constr.
– reference: Zuo, X., Dai, B., Shan, Y., Shen, J., Hu, C., Huang, S., 2020. Classifying cracks at sub-class level in closed circuit television sewer inspection videos. Autom. Constr. 118.
– volume: 47
  year: 2021
  ident: b0035
  article-title: Automatic classification of wall and door BIM element subtypes using 3D geometric deep neural networks
  publication-title: Adv. Eng. Inf.
– reference: Sun, J., Peng, B., Wang, C.C., Chen, K., Zhong, B., Wu, J., 2022. Building displacement measurement and analysis based on UAV images. Autom. Constr. 140.
– reference: Yang, X., del Rey Castillo, E., Zou, Y., Wotherspoon, L., Tan, Y., 2022. Automated semantic segmentation of bridge components from large-scale point clouds using a weighted superpoint graph. Autom. Constr. 142.
– reference: Zhao, S., Kang, F., Li, J., 2022. Concrete dam damage detection and localisation based on YOLOv5s-HSC and photogrammetric 3D reconstruction. Autom. Constr. 143.
– reference: Ma, D., Fang, H., Wang, N., Lu, H., Matthews, J., Zhang, C., 2023. Transformer‐optimized generation, detection, and tracking network for images with drainage pipeline defects. Comput.-Aided Civil Infrastruct. Eng.
– volume: 12
  start-page: 213
  year: 2022
  ident: b0105
  article-title: Study of damage quantification of concrete drainage pipes based on point cloud segmentation and reconstruction
  publication-title: Buildings
– reference: Meng, S., Gao, Z., Zhou, Y., He, B., Djerrad, A., 2022. Real‐time automatic crack detection method based on drone. Comput.-Aided Civil Infrastruct. Eng.
– volume: 119
  start-page: 1
  year: 2020
  end-page: 12
  ident: b0100
  article-title: Automatic sewer pipe defect semantic segmentation based on improved U-Net
  publication-title: Autom. Constr.
– volume: 11
  start-page: 4270
  year: 2018
  end-page: 4286
  ident: b0170
  article-title: Unsupervised Segmentation of Point Clouds From Buildings Using Hierarchical Clustering Based on Gestalt Principles
  publication-title: IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens.
– volume: 239
  start-page: 109
  year: 2018
  end-page: 118
  ident: b0200
  article-title: Automated method for extracting and analysing the rock discontinuities from point clouds based on digital surface model of rock mass
  publication-title: Eng. Geol.
– volume: 37
  start-page: 163
  year: 2018
  end-page: 174
  ident: b0070
  article-title: A review of 3D reconstruction techniques in civil engineering and their applications
  publication-title: Adv. Eng. Inf.
– volume: 312
  start-page: 1
  year: 2021
  end-page: 18
  ident: b0075
  article-title: A multi-defect detection system for sewer pipelines based on StyleGAN-SDM and fusion CNN
  publication-title: Constr. Build. Mater.
– volume: 18
  start-page: 263
  year: 2013
  end-page: 276
  ident: b0150
  article-title: Classification of defects with ensemble methods in the automated visual inspection of sewer pipes
  publication-title: Pattern Anal. Appl.
– reference: He, K., Zhang, X., Ren, S., Sun, J., 2016. Deep residual learning for image recognition. In: Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, pp. 770-778.
– volume: 8
  start-page: 239
  year: 2021
  end-page: 250
  ident: b0040
  article-title: A geometric deep learning approach for checking element-to-entity mappings in infrastructure building information models
  publication-title: J. Comput. Des. Eng.
– reference: Yin, C., Wang, B., Gan, V.J.L., Wang, M., Cheng, J.C.P., 2021. Automated semantic segmentation of industrial point clouds using ResPointNet++. Autom. Constr. 130.
– reference: Moeslund, T., Nikolov, I., Henriksen, K., Lynge, M., Allahham, M., Haurum, J., 2021. Sewer defect classification using synthetic point clouds. In: Proceedings of the 16th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications, pp. 891-900.
– reference: United States Department of Transportation, Pipeline Incident 20 Year Trends, https://www.phmsa.dot.gov/data-and-statistics/pipeline/pipeline-incident-20-year-trends.
– volume: 12
  year: 2022
  ident: b0025
  article-title: Research on 3D Defect Information Management of Drainage Pipeline Based on BIM
  publication-title: Buildings
– reference: Meijer, D., Luimes, R., Knobbe, A., Bäck, T., 2022. Anomaly detection in urban drainage with stereovision. Autom. Constr. 139.
– volume: 13
  year: 2021
  ident: b0095
  article-title: Building Component Detection on Unstructured 3D Indoor Point Clouds Using RANSAC-Based Region Growing
  publication-title: Remote Sens. (Basel)
– reference: A.C. Wilson R. Roelofs M. Stern N. Srebro B. Recht The Marginal Value of Adaptive Gradient Methods in Machine Learning 2017.
– volume: 37
  start-page: 354
  year: 2021
  end-page: 369
  ident: b0165
  article-title: A 3D reconstruction method for buildings based on monocular vision. Comput.-Aided Civil Infrastruct
  publication-title: Eng.
– volume: 91
  start-page: 273
  year: 2018
  end-page: 283
  ident: b0045
  article-title: Automated defect classification in sewer closed circuit television inspections using deep convolutional neural networks
  publication-title: Autom. Constr.
– volume: 75
  start-page: 562
  year: 2019
  end-page: 574
  ident: b0110
  article-title: 3D pipe reconstruction employing video information from mobile robots
  publication-title: Appl. Soft Comput.
– volume: 7
  start-page: 311
  year: 2022
  end-page: 323
  ident: b0175
  article-title: 3D reconstruction and automatic leakage defect quantification of metro tunnel based on SfM-Deep learning method
  publication-title: Underground Space
– reference: Xia, T., Yang, J., Chen, L., 2022. Automated semantic segmentation of bridge point cloud based on local descriptor and machine learning. Autom. Constr. 133.
– volume: 11
  start-page: 165
  year: 2005
  end-page: 171
  ident: b0115
  article-title: Automated detection and classification of infiltration in sewer pipes
  publication-title: J. Infrastruct. Syst.
– reference: Insa-Iglesias, M., Jenkins, M.D., Morison, G., 2021. 3D visual inspection system framework for structural condition monitoring and analysis. Autom. Constr. 128.
– reference: Ma, J.W., Leite, F., 2022. Performance boosting of conventional deep learning-based semantic segmentation leveraging unsupervised clustering. Autom. Constr. 136.
– volume: 86
  start-page: 81
  year: 2018
  end-page: 98
  ident: b0190
  article-title: Optimal placement of precast bridge deck slabs with respect to precast girders using 3D laser scanning
  publication-title: Autom. Constr.
– year: 2021
  ident: b0210
  article-title: Study on leak localization for buried gas pipelines based on an acoustic method
  publication-title: Tunn. Undergr. Space Technol.
– volume: 12
  year: 2022
  ident: 10.1016/j.tust.2023.105480_b0025
  article-title: Research on 3D Defect Information Management of Drainage Pipeline Based on BIM
  publication-title: Buildings
– ident: 10.1016/j.tust.2023.105480_b0185
  doi: 10.1016/j.autcon.2021.103874
– ident: 10.1016/j.tust.2023.105480_b0065
  doi: 10.1111/mice.12970
– ident: 10.1016/j.tust.2023.105480_b0120
  doi: 10.1016/j.autcon.2022.104367
– ident: 10.1016/j.tust.2023.105480_b0055
  doi: 10.1016/j.tust.2022.104861
– volume: 23
  start-page: 4685
  year: 2022
  ident: 10.1016/j.tust.2023.105480_b0005
  article-title: Pothole 3D reconstruction with a novel imaging system and structure from motion techniques
  publication-title: IEEE Trans. Intell. Transp. Syst.
  doi: 10.1109/TITS.2021.3054026
– ident: 10.1016/j.tust.2023.105480_b0215
  doi: 10.1016/j.autcon.2022.104555
– volume: 11
  start-page: 4270
  year: 2018
  ident: 10.1016/j.tust.2023.105480_b0170
  article-title: Unsupervised Segmentation of Point Clouds From Buildings Using Hierarchical Clustering Based on Gestalt Principles
  publication-title: IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens.
  doi: 10.1109/JSTARS.2018.2817227
– ident: 10.1016/j.tust.2023.105480_b0125
– volume: 11
  start-page: 165
  year: 2005
  ident: 10.1016/j.tust.2023.105480_b0115
  article-title: Automated detection and classification of infiltration in sewer pipes
  publication-title: J. Infrastruct. Syst.
  doi: 10.1061/(ASCE)1076-0342(2005)11:3(165)
– volume: 51
  start-page: 32
  year: 2015
  ident: 10.1016/j.tust.2023.105480_b0010
  article-title: Segmentation of building point cloud models including detailed architectural/structural features and MEP systems
  publication-title: Autom. Constr.
  doi: 10.1016/j.autcon.2014.12.015
– ident: 10.1016/j.tust.2023.105480_b0015
  doi: 10.1016/j.autcon.2022.104494
– ident: 10.1016/j.tust.2023.105480_b0030
  doi: 10.1016/j.autcon.2021.103755
– volume: 12
  start-page: 213
  year: 2022
  ident: 10.1016/j.tust.2023.105480_b0105
  article-title: Study of damage quantification of concrete drainage pipes based on point cloud segmentation and reconstruction
  publication-title: Buildings
  doi: 10.3390/buildings12020213
– ident: 10.1016/j.tust.2023.105480_b0080
  doi: 10.1016/j.autcon.2022.104285
– volume: 123
  start-page: 1
  year: 2022
  ident: 10.1016/j.tust.2023.105480_b0225
  article-title: Automatic sewer defect detection and severity quantification based on pixel-level semantic segmentation
  publication-title: Tunn. Undergr. Space Technol.
  doi: 10.1016/j.tust.2022.104403
– volume: 110
  start-page: 1
  year: 2021
  ident: 10.1016/j.tust.2023.105480_b0135
  article-title: Towards an automated condition assessment framework of underground sewer pipes based on closed-circuit television (CCTV) images
  publication-title: Tunn. Undergr. Space Technol.
  doi: 10.1016/j.tust.2021.103840
– ident: 10.1016/j.tust.2023.105480_b0020
  doi: 10.1109/CVPR.2016.90
– volume: 47
  year: 2021
  ident: 10.1016/j.tust.2023.105480_b0035
  article-title: Automatic classification of wall and door BIM element subtypes using 3D geometric deep neural networks
  publication-title: Adv. Eng. Inf.
  doi: 10.1016/j.aei.2020.101200
– volume: 37
  start-page: 354
  year: 2021
  ident: 10.1016/j.tust.2023.105480_b0165
  article-title: A 3D reconstruction method for buildings based on monocular vision. Comput.-Aided Civil Infrastruct
  publication-title: Eng.
– ident: 10.1016/j.tust.2023.105480_b0180
  doi: 10.1016/j.autcon.2022.104519
– volume: 86
  start-page: 81
  year: 2018
  ident: 10.1016/j.tust.2023.105480_b0190
  article-title: Optimal placement of precast bridge deck slabs with respect to precast girders using 3D laser scanning
  publication-title: Autom. Constr.
  doi: 10.1016/j.autcon.2017.11.004
– volume: 7
  start-page: 311
  year: 2022
  ident: 10.1016/j.tust.2023.105480_b0175
  article-title: 3D reconstruction and automatic leakage defect quantification of metro tunnel based on SfM-Deep learning method
  publication-title: Underground Space
  doi: 10.1016/j.undsp.2021.08.004
– volume: 239
  start-page: 109
  year: 2018
  ident: 10.1016/j.tust.2023.105480_b0200
  article-title: Automated method for extracting and analysing the rock discontinuities from point clouds based on digital surface model of rock mass
  publication-title: Eng. Geol.
  doi: 10.1016/j.enggeo.2018.03.020
– volume: 8
  start-page: 38
  year: 2020
  ident: 10.1016/j.tust.2023.105480_b0160
  article-title: Linking Points With Labels in 3D: A Review of Point Cloud Semantic Segmentation
  publication-title: IEEE Geosci. Remote Sens. Mag.
  doi: 10.1109/MGRS.2019.2937630
– volume: 8
  start-page: 239
  year: 2021
  ident: 10.1016/j.tust.2023.105480_b0040
  article-title: A geometric deep learning approach for checking element-to-entity mappings in infrastructure building information models
  publication-title: J. Comput. Des. Eng.
– ident: 10.1016/j.tust.2023.105480_b0140
– ident: 10.1016/j.tust.2023.105480_b0060
  doi: 10.1016/j.autcon.2022.104167
– ident: 10.1016/j.tust.2023.105480_b0090
  doi: 10.5220/0010207908910900
– volume: 133
  year: 2023
  ident: 10.1016/j.tust.2023.105480_b0195
  article-title: Axial stresses in pressure pipe liners spanning joints with initial gap, opening as a result of differential ground movements
  publication-title: Tunn. Undergr. Space Technol.
  doi: 10.1016/j.tust.2022.104965
– volume: 75
  start-page: 562
  year: 2019
  ident: 10.1016/j.tust.2023.105480_b0110
  article-title: 3D pipe reconstruction employing video information from mobile robots
  publication-title: Appl. Soft Comput.
  doi: 10.1016/j.asoc.2018.11.016
– volume: 119
  start-page: 1
  year: 2020
  ident: 10.1016/j.tust.2023.105480_b0100
  article-title: Automatic sewer pipe defect semantic segmentation based on improved U-Net
  publication-title: Autom. Constr.
  doi: 10.1016/j.autcon.2020.103383
– volume: 7
  start-page: 259
  year: 1998
  ident: 10.1016/j.tust.2023.105480_b0145
  article-title: Assessment technologies for sewer system rehabilitation
  publication-title: Autom. Constr.
  doi: 10.1016/S0926-5805(97)00071-X
– ident: 10.1016/j.tust.2023.105480_b0085
  doi: 10.1111/mice.12918
– volume: 11
  start-page: 2101
  year: 2019
  ident: 10.1016/j.tust.2023.105480_b0205
  article-title: A 3D reconstruction pipeline of urban drainage pipes based on multiviewImage matching using low-cost panoramic video cameras
  publication-title: Water
  doi: 10.3390/w11102101
– ident: 10.1016/j.tust.2023.105480_b0230
  doi: 10.1016/j.autcon.2020.103289
– volume: 18
  start-page: 263
  year: 2013
  ident: 10.1016/j.tust.2023.105480_b0150
  article-title: Classification of defects with ensemble methods in the automated visual inspection of sewer pipes
  publication-title: Pattern Anal. Appl.
  doi: 10.1007/s10044-013-0355-5
– ident: 10.1016/j.tust.2023.105480_b0155
  doi: 10.1016/j.autcon.2021.103992
– volume: 130
  year: 2022
  ident: 10.1016/j.tust.2023.105480_b0050
  article-title: Impact of erosion voids and internal corrosion on concrete pipes under traffic loads
  publication-title: Tunn. Undergr. Space Technol.
  doi: 10.1016/j.tust.2022.104761
– volume: 13
  year: 2021
  ident: 10.1016/j.tust.2023.105480_b0095
  article-title: Building Component Detection on Unstructured 3D Indoor Point Clouds Using RANSAC-Based Region Growing
  publication-title: Remote Sens. (Basel)
– volume: 91
  start-page: 273
  year: 2018
  ident: 10.1016/j.tust.2023.105480_b0045
  article-title: Automated defect classification in sewer closed circuit television inspections using deep convolutional neural networks
  publication-title: Autom. Constr.
  doi: 10.1016/j.autcon.2018.03.028
– volume: 37
  start-page: 163
  year: 2018
  ident: 10.1016/j.tust.2023.105480_b0070
  article-title: A review of 3D reconstruction techniques in civil engineering and their applications
  publication-title: Adv. Eng. Inf.
  doi: 10.1016/j.aei.2018.05.005
– year: 2021
  ident: 10.1016/j.tust.2023.105480_b0210
  article-title: Study on leak localization for buried gas pipelines based on an acoustic method
  publication-title: Tunn. Undergr. Space Technol.
– volume: 35
  start-page: 162
  year: 2020
  ident: 10.1016/j.tust.2023.105480_b0130
  article-title: A unified convolutional neural network integrated with conditional random field for pipe defect segmentation. Comput.-Aided Civil Infrastruct
  publication-title: Eng.
– ident: 10.1016/j.tust.2023.105480_b0220
  doi: 10.1016/j.autcon.2022.104163
– volume: 312
  start-page: 1
  year: 2021
  ident: 10.1016/j.tust.2023.105480_b0075
  article-title: A multi-defect detection system for sewer pipelines based on StyleGAN-SDM and fusion CNN
  publication-title: Constr. Build. Mater.
  doi: 10.1016/j.conbuildmat.2021.125385
SSID ssj0005229
Score 2.4945118
Snippet •An 3D inspection method on point clouds for pipe defects was proposed.•The network structure was optimized to improve the inspection accuracy.•The training...
SourceID crossref
elsevier
SourceType Enrichment Source
Index Database
Publisher
StartPage 105480
SubjectTerms 3D segmentation
Deep learning
Defect inspection
Point cloud
Sewer pipeline
Title An automatic defect classification and segmentation method on three-dimensional point clouds for sewer pipes
URI https://dx.doi.org/10.1016/j.tust.2023.105480
Volume 143
WOSCitedRecordID wos001100271500001&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
hasFullText 1
inHoldings 1
isFullTextHit
isPrint
journalDatabaseRights – providerCode: PRVESC
  databaseName: ScienceDirect Freedom Collection
  customDbUrl:
  eissn: 1878-4364
  dateEnd: 99991231
  omitProxy: false
  ssIdentifier: ssj0005229
  issn: 0886-7798
  databaseCode: AIEXJ
  dateStart: 19950101
  isFulltext: true
  titleUrlDefault: https://www.sciencedirect.com
  providerName: Elsevier
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwtV1Nb9QwELWWlgMcEJ-i5UM-cFulyqbeJD4uqAhQWSGxoNyiJLarrRYnWpKqv6C_mxmPk122VQVIXKLIiuPI8zJ5dmbeMPZGiYlO06IMjE6ngZiUYSB1UoJBEjM1okyNKwbz_TSZz9Msk19Go6s-F-ZilVibXl7K5r-aGtrA2Jg6-xfmHm4KDXAORocjmB2Of2T4mR0XXVuTFKvSTpy4Qo6MQUFkbrdZrs9--Lwj6-tI04-DtdaBQs1_0usYN_XS4h3qTjntBuiJgfXNsvHhh57aLjoXMtMnPWJy2hpzRnAsWJjrcXvDLj55mvkSKydtcPqZfkJ19YZmY0PWoRLKWd946gIR3nrpcL9zEYmdnYshpWYTv0ReLwbKT6WpjzR55RSWuuKY5M4Ht03yTtc-AbQbcX7UYsoKFofHUsaCykXtSGt_xcFwrAhF-8Mwu8P2o2QqwTvuzz6eZJ-2goVcvbvh4Xz6FUUK7o50M8XZoi2Lh-yBX2_wGeHkERtp-5jd31KhfMJWM8sHxHBCDP8dMRwsyrcRwwkxHM6uIYY7xHBCDAfEcIcY7hDzlH17f7J49yHwRTiCCialDZJYRoWIS3ihj0UcVgUSHIMZyVIZWcSRxoI1sRJFWQJ5jEJjhFGhkSaElx6M9Izt2drq54xXKF-gIqEKKQWsU0spwR8A5y8iM5GVPmCTft7yyivUY6GUVd6HIp7nONc5znVOc33AxkOfhvRZbr162psj9wyTmGMO6Lml3-E_9nvB7m2A_5LttetOv2J3q4t2-XP92oPsF-pppSw
linkProvider Elsevier
openUrl ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=An+automatic+defect+classification+and+segmentation+method+on+three-dimensional+point+clouds+for+sewer+pipes&rft.jtitle=Tunnelling+and+underground+space+technology&rft.au=Wang%2C+Niannian&rft.au=Ma%2C+Duo&rft.au=Du%2C+Xueming&rft.au=Li%2C+Bin&rft.date=2024-01-01&rft.pub=Elsevier+Ltd&rft.issn=0886-7798&rft.eissn=1878-4364&rft.volume=143&rft_id=info:doi/10.1016%2Fj.tust.2023.105480&rft.externalDocID=S088677982300500X
thumbnail_l http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=0886-7798&client=summon
thumbnail_m http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=0886-7798&client=summon
thumbnail_s http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=0886-7798&client=summon