Toward Vision-Based Concrete Crack Detection: Automatic Simulation of Real-World Cracks

Tran Hiep Dinh1,2, Vu Thi Thuy Anh3, TruongGiang Nguyen4, Cong Hieu Le5, Nguyen Linh Trung5, Nguyen Dinh Duc3, Chin-Teng Lin6
1Faculty of Engineering Mechanics and Automation, VNU University of Engineering and Technology, Hanoi, Vietnam
2Joint Technology and Innovation Research Centre (JTIRC), a partnership between the University of Technology Sydney, Ultimo, NSW, Australia
3Faculty of Civil Engineering, VNU University of Engineering and Technology, Hanoi, Vietnam
4Institute of Mechanics, Vietnam National Academy of Science and Technology, Hanoi, Vietnam
5Advanced Institute of Engineering and Technology (AVITECH), VNU University of Engineering and Technology, Hanoi, Vietnam
6Australian Artificial Intelligence Institute, School of Computer Science, Faculty of Engineering and Information Technology, University of Technology Sydney, Ultimo, NSW, Australia

Tóm tắt

Vision-based concrete crack detection has recently attracted significant attention from many researchers. Although promising results have been obtained, especially for deep learning (DL) approaches, it is difficult to maintain the robustness of implemented models when tested on completely new data. A possible reason for this is that the extracted feature from the trained set might not fully characterize the crack in the test set. We propose an interdisciplinary approach to improve the effectiveness of vision-based crack detection by modeling crack propagation using fracture mechanics, simulation, and machine learning (ML). Mathematical models of concrete cracks are obtained using ML on the simulation results. Experiments are conducted on various reputable crack image datasets, emphasizing the correlation between simulated and real-world cracks. The importance of propagation models is verified in a classification task, reporting a significant accuracy enhancement on results of some state-of-the-art detection and segmentation models, i.e., 1.27% on average on participating models, and 5.47% on U-Net. This novel approach is expected to have valuable points for a research area where the data quantity and quality still need to be improved.

Từ khóa

#Concrete beam #crack dataset #crack detection #crack propagation #extended finite element method (X-FEM) #machine learning (ML) #regression model #vision-based inspection

Tài liệu tham khảo

10.1109/ICCV.2015.164 10.1109/TIP.2018.2878966 10.1016/j.imavis.2011.10.003 10.1016/j.autcon.2019.04.005 10.1016/j.autcon.2018.11.028 kingma, 2015, Adam: A method for stochastic optimization, Proc Int Conf Learn Represent (ICLR), 1 10.1109/ICCV.2015.123 10.1109/TITS.2019.2910595 10.1016/j.patrec.2011.11.004 10.1109/TIM.2023.3272052 10.1109/TPAMI.2018.2815688 10.1109/TIM.2022.3178991 10.3390/s23010053 10.1016/j.autcon.2022.104440 10.1177/14759217221140976 10.1109/CVPR.2005.177 10.1109/IJCNN.2017.7966101 10.1023/A:1023399125413 karihaloo, 1995, Fracture mechanics & structural concrete, Longman Scientific and Technical groß, 2007, An extended pressure finite element space for two-phase incompressible flows with surface tension, J Comput Phys, 224, 40, 10.1016/j.jcp.2006.12.021 10.1007/BF02472030 10.1061/(ASCE)GM.1943-5622.0000947 10.1016/j.engfracmech.2020.107193 10.1016/j.advengsoft.2017.01.006 fan, 2013, Numerical manifold method (NMM) simulation of stress wave propagation through fractured rock mass, International Applied Mechanics, 5 10.1002/1097-0207(20010210)50:4<993::AID-NME164>3.0.CO;2-M 10.1109/ACCESS.2018.2855144 10.1109/TITS.2016.2552248 10.1109/TIM.2021.3075022 10.1109/TIP.2019.2932501 10.1109/TITS.2018.2856928 10.1109/ACCESS.2018.2829347 10.1109/TSMC.2018.2827937 10.1007/s00603-019-01914-5 10.1142/S0219455422500894 10.1016/j.tafmec.2023.103906 10.1002/1097-0207(20010228)50:6<1291::AID-NME12>3.0.CO;2-Q 10.1016/j.ijsolstr.2023.112269 10.1016/S0008-8846(02)00825-6 phuc, 2021, Free vibration of cracked FGM plates with variable thickness resting on elastic foundations, Thin-Walled Struct, 161 10.1016/j.euromechsol.2023.104983 10.1007/978-94-009-6152-4_4 10.1016/j.engappai.2022.105478 10.1016/j.autcon.2021.103989 shi, 1992, Manifold method of material analysis 10.1016/j.engstruct.2009.02.041 10.1016/j.aei.2021.101303 10.1109/ICARCV.2016.7838682 10.1016/j.engfracmech.2022.108533 10.1016/j.autcon.2019.04.014 10.1109/TIM.2023.3307181 10.1201/9781315370293 10.1109/TITS.2021.3138428 10.1590/s1983-41952022000400006 10.1016/j.pedc.2022.100027 10.1016/j.engfracmech.2018.06.006