Semantic segmentation of cracks: Data challenges and architecture

被引:0
|
作者
Panella, Fabio [1 ]
Lipani, Aldo [1 ]
Boehm, Jan [1 ]
机构
[1] Dept. of Civil, Environ. and Geomatic Eng., UCL, London,UK, United Kingdom
基金
英国工程与自然科学研究理事会;
关键词
Deep learning - Crack detection - Network layers - Semantics - Network architecture - Architecture;
D O I
暂无
中图分类号
学科分类号
摘要
Deep Learning (DL) semantic image segmentation is a technique used in several fields of research. The present paper analyses semantic crack segmentation as a case study to review the up to date research on semantic segmentation in the presence of fine structures and the effectiveness of established approaches to address the inherent class imbalance issue. The established UNet architecture is tested against networks consisting exclusively of stacked convolution without pooling layers (straight networks), with regard to the resolution of their segmentation results. Dice and Focal losses are also compared against each other to evaluate their effectiveness on highly imbalanced data. With the same aim, dropout and data augmentation approaches are tested, as additional regularizing mechanisms, to address the uneven distribution of the dataset. The experiments show that the good selection of the loss function has more impact in handling the class imbalance and boosting the detection performance than all the other regularizers with regards to segmentation resolution. Moreover, UNet, the architecture considered as reference, clearly outperforms the networks with no pooling layers both in performance and training time. The authors argue that UNet architectures, compared to the networks with no pooling layers, achieve high detection performance at a very low cost in terms of training time. Therefore, the authors consider such architecture as the state of the art for semantic segmentation of cracks. On the other hand, once computational cost is not an issue anymore thanks to constant improvements of technology, the application of networks without pooling layers might become attractive again because of their simplicity of and high performance. © 2022 Elsevier B.V.
引用
收藏
相关论文
共 50 条
  • [31] Fast detection algorithm for cracks on tunnel linings based on deep semantic segmentation
    Zhou, Zhong
    Zheng, Yidi
    Zhang, Junjie
    Yang, Hao
    FRONTIERS OF STRUCTURAL AND CIVIL ENGINEERING, 2023, 17 (05) : 732 - 744
  • [32] Fast detection algorithm for cracks on tunnel linings based on deep semantic segmentation
    Zhong Zhou
    Yidi Zheng
    Junjie Zhang
    Hao Yang
    Frontiers of Structural and Civil Engineering, 2023, 17 : 732 - 744
  • [33] Generative AI for Semantic Communication: Architecture, Challenges, and Outlook
    Xia, Le
    Sun, Yao
    Liang, Chengsi
    Zhang, Lei
    Imran, Muhammad Ali
    Niyato, Dusit
    IEEE WIRELESS COMMUNICATIONS, 2025, 32 (01) : 132 - 140
  • [34] Semantic Segmentation of Cracks on Masonry Surfaces Using Deep-Learning Techniques
    Patel, Sudhir Babu
    Bisht, Pranjal
    Pathak, Krishna Kant
    PRACTICE PERIODICAL ON STRUCTURAL DESIGN AND CONSTRUCTION, 2024, 29 (02)
  • [35] A classification of challenges in the Semantic Web based on the general architecture
    Martinez-Rodriguez, Jose L.
    Lopez-Arevalo, Ivan
    Rios-Alvarado, Ana B.
    2015 26TH INTERNATIONAL WORKSHOP ON DATABASE AND EXPERT SYSTEMS APPLICATIONS (DEXA), 2015, : 197 - 201
  • [36] Semantic segmentation for tooth cracks using improved DeepLabv3+model
    Xie, Zewen
    Lu, Qilin
    Guo, Juncheng
    Lin, Weiren
    Ge, Guanghua
    Tang, Yadong
    Pasini, Damiano
    Wang, Wenlong
    HELIYON, 2024, 10 (04)
  • [37] Understanding the Challenges When 3D Semantic Segmentation Faces Class Imbalanced and OOD Data
    Pan, Yancheng
    Xie, Fan
    Zhao, Huijing
    IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS, 2023, 24 (07) : 6955 - 6970
  • [38] Buffer ladder feature fusion architecture for semantic segmentation improvement
    Liu, Zonghui
    Lei, Zhichun
    SIGNAL IMAGE AND VIDEO PROCESSING, 2024, 18 (01) : 475 - 483
  • [39] SqueezeNAS: Fast Neural Architecture Search for Faster Semantic Segmentation
    Shaw, Albert
    Hunter, Daniel
    Iandola, Forrest
    Sidhu, Sammy
    2019 IEEE/CVF INTERNATIONAL CONFERENCE ON COMPUTER VISION WORKSHOPS (ICCVW), 2019, : 2014 - 2024
  • [40] Wound Semantic Segmentation Framework with Encoder- Decoder Architecture
    Jiang, Wenda
    Gao, Zhen
    Brahim, Wael
    2024 IEEE 4TH INTERNATIONAL CONFERENCE ON SOFTWARE ENGINEERING AND ARTIFICIAL INTELLIGENCE, SEAI 2024, 2024, : 6 - 10