Title
A deep learning framework based on improved self-supervised learning for ground-penetrating radar tunnel lining inspection
Author
Huang, Jian (China University of Geosciences, Wuhan)
Yang, Xi (China University of Geosciences, Wuhan)
Zhou, Feng (China University of Geosciences, Wuhan; Southern University of Science and Technology)
Li, Xiaofeng (China University of Geosciences, Wuhan)
Zhou, Bin (China Railway Southwest Research Institute Co. Ltd.)
Lu, Song (China Railway Southwest Research Institute Co. Ltd.)
Ivashov, Sergey (Bauman Moscow State Technical University)
Giannakis, Iraklis (University of Aberdeen)
Kong, Fannian (Norwegian Geotechnical Institute)
Slob, E.C. (TU Delft Applied Geophysics and Petrophysics) ![ORCID 0000-0002-4529-1134 ORCID 0000-0002-4529-1134](/sites/all/themes/tud_repo3/img/icons/orcid_16x16.png)
Date
2023
Abstract
It is not practical to obtain a large number of labeled data to train a supervised learning network in tunnel lining nondestructive testing with ground-penetrating radar (GPR). To decrease the dependence of supervised learning on the number of labeled data, an improved self-supervised learning algorithm—self-attention dense contrastive learning (SA-DenseCL)—is proposed and incorporated with a mask region-convolution neural network (Mask R-CNN), which is trained by unlabeled and labeled GPR data. The proposed SA-DenseCL adds a self-attention-based relevant projection head to the DenseCL architecture of self-supervised learning, capturing the spatially continuing information between adjacent GPR traces. In the workflow, some unlabeled GPR images are used to pre-train the SA-DenseCL network for feature extraction and obtaining the backbone weights, which is superior to the conventional pre-training methods of supervised learning pre-trained by ImageNet images. The weights of the pre-trained backbone are then used to initialize the Mask R-CNN through transfer learning. Subsequently, a limited number of labeled GPR images are used to fine-tune the Mask R-CNN for automatically identifying the locations of the reinforcement bars and voids and estimating the secondary lining thickness. The experimental results show that the average precision reaches 96.70%, 81.04%, and 94.67% in identifying reinforcement bar locations, detecting void defects, and estimating secondary lining thickness, respectively, which outperform the conventional methods that use ImageNet-based supervised learning or GPR image-based DenseCL for initializing the Mask R-CNN backbone weights. It is observed that the improved self-supervised learning-based framework can improve the detection and estimation accuracy in GPR tunnel lining inspection.
To reference this document use:
http://resolver.tudelft.nl/uuid:345a4c40-612c-451e-8203-f8e104fd79fb
DOI
https://doi.org/10.1111/mice.13042
Embargo date
2023-11-19
ISSN
1093-9687
Source
Computer-Aided Civil and Infrastructure Engineering, 39 (6), 814-833
Bibliographical note
Green Open Access added to TU Delft Institutional Repository ‘You share, we take care!’ – Taverne project https://www.openaccess.nl/en/you-share-we-take-care Otherwise as indicated in the copyright section: the publisher is the copyright holder of this work and the author uses the Dutch legislation to make this work public.
Part of collection
Institutional Repository
Document type
journal article
Rights
© 2023 Jian Huang, Xi Yang, Feng Zhou, Xiaofeng Li, Bin Zhou, Song Lu, Sergey Ivashov, Iraklis Giannakis, Fannian Kong, E.C. Slob