Micro-concrete crack detection of underwater structures based on convolutional neural network

Qi, Zhilong; Liu, Donghai; Chen, Junjie

doi:10.1007/s00138-022-01327-5

Cited by 17 publications

(4 citation statements)

References 34 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Consequently, they have poor prediction performance and limited applicability. By contrast, DL models can adaptively adjust parameters, mine hidden features [33], and enhance prediction accuracy. Cao et al [34] proposed a Recurrent Neural Network-Gappy Proper Orthogonal Decomposition (RNN-GPOD) model for tunnel surface settlement prediction, aiding accurate tunnel boring machine operation.…”

Section: Related Workmentioning

confidence: 99%

Improved Machine Learning Model for Urban Tunnel Settlement Prediction Using Sparse Data

Yu,

Jin,

et al. 2024

Sustainability

View full text Add to dashboard Cite

Prediction tunnel settlement in shield tunnels during the operation period has gained increasing significance within the realm of maintenance strategy formulation. The sparse settlement data during this period present a formidable challenge for predictive Artificial Intelligence (AI) models, as they may not handle non-stationary relationships effectively or have the risk of overfitting. In this study, we propose an improved machine learning (ML) model based on sparse settlement data. We enhance training data via time series clustering, use time decomposition to uncover latent features, and employ Extreme Gradient Boosting (XGBoost) v1.5.1 with Bayesian Optimization (BO) v1.2.0 for precise predictions. Comparative experiments conducted on different acquisition points substantiate our model’s efficacy, the in-training set yielding a Mean Absolute Error (MAE) of 0.649 mm, Root Mean Square Error (RMSE) of 0.873 mm, Mean Absolute Percentage Error (MAPE) of 3.566, and Coefficient of Determination (R2) of 0.872, and the in-testing set yielding a MAE of 0.717 mm, RMSE of 1.048 mm, MAPE of 4.080, and R2 of 0.846. The empirical results show the superiority of the proposed model compared to simple ML models and a complex neural network model, as it has a lower prediction error and higher accuracy across different sparse settlement datasets. Moreover, this paper underlines that accurate settlement predictions contribute to achieving some Sustainable Development Goals (SDGs). Specifically, preventive tunnel maintenance strategies based on predictive results can enhance tunnels’ long-term operational reliability, which is in accordance with SDG 9 (Industry, Innovation, and Infrastructure) and SDG 11 (Sustainable Cities and Communities).

show abstract

Section: Related Workmentioning

confidence: 99%

Improved Machine Learning Model for Urban Tunnel Settlement Prediction Using Sparse Data

Yu,

Jin,

et al. 2024

Sustainability

View full text Add to dashboard Cite

show abstract

“…They mimic the clairvoyance and omniscient of humans, allowing them to sense facility defects that are distant in space and time. Last but not least, AI augments the cognitive power of humans, making it possible to process the vast amount of facility data collected by robot-borne sensory devices [23].…”

Section: From Sporadic Inspection To Systematic Mappingmentioning

confidence: 99%

Built environment defect mapping, modeling, and management (D3M): A BIM-based integrated framework

Chen,

Lu,

Liu

2024

Journal of Intelligent Construction

Self Cite

View full text Add to dashboard Cite

Keywordsbuilt environment sustainability facility management defect information modeling building information modelingThe built environment is subject to various defects as it ages. A well-maintained built environment depends on surveying activities to inspect, document, and rehabilitate the defects that occurred. The advancement of digital technologies paves the pathway towards (1) comprehensive defect inspection by systematic mapping, (2) their consistent documentation by digital modeling, and (3) timely retrofitting by proactive management. However, the three steps of defect mapping, modeling, and management (D3M) remain largely fragmented and have yet to be synergized. Exploiting the pivotal role of building information modeling (BIM) in built asset management, this paper puts forward a cohesive framework for integrated D3M. It leverages the rich geometric-semantic information in BIM to assist defect mapping and enriches the BIM by industry foundation classes (IFCs)-represented defect information. The defect-enriched BIM facilitates defect management in a data-driven manner. The framework was applied in multiple real-life infrastructure and civil works projects. It demonstrates how the BIM-based D3M framework can enhance the maintenance of those that have been built, and ultimately contribute to a safe and sustainable built environment. Future studies are called for to substantiate each of the 3Ms by leveraging BIM as both an enabler and a beneficiary.

show abstract

“…In terms of the complex industrial production environment and defects of variable size, Zhang et al [18] proposed a lightweight defect detection network based on a dual-attention mechanism and a path aggregation feature pyramid network to achieve better detection results than other networks on PLC circuit boards. Qi [19]proposed the method of CNN-based for micro-concrete crack detection of underwater structures. Tang [20] designed an inspection framework based on convolutional autoencoder(CAE) to inspect core failures from real-world die casting X-ray images in an unsupervised mannerWang et al However, most lightweight networks have simplified the parameters, but the feature extraction capability of the model has much to be improved.…”

Section: Lightweight Networkmentioning

confidence: 99%

Light-Trans YOLO: A Lightweight Network Based on Transformer for Defect Detection of Industrial Lace Surface

Wang,

Jing,

Sheng

2023

Preprint

View full text Add to dashboard Cite

Lace surface Defect detection has always been a crucial step in the industrial production of lace products. However, due to the complex texture and deforma-bility of lace, as well as the difficulty of distinguishing minor defects from normal images. Therefore, the detection of defects on lace surfaces is a challenging but rarely studied task. In this paper, we propose a new lightweight detection framework , Light-Trans YOLO, to detect lace surface defects. First, our backbone network uses the lightweight network C3 GhostNet. In addition, to obtain more complete global information, we add the lightweight Mobile Transformer Block (MTB) to the backbone network. Then we use the proposed standard deep-wise separable convolution (SDSConv) and SDSBottleneck to design a new neck and add Coordinate Attention (CA) at the end, which overcomes the problem of information loss of deep separable convolution and extracts more effective information. In the training, we propose the loss function ϵSIoU, which can improve the separability of defective and normal samples. We conduct experiments on the industrial lace surface defect dataset collected in lace production sites, and the experiments prove that the mAP of our model is 96.6%, which is 7.7% higher than YOLOV5s, and the FPS and F1-score of the model reaches 50.3 and 0.93, which indicates that our model has a great trade-off between detection accuracy and speed.

show abstract

Micro-concrete crack detection of underwater structures based on convolutional neural network

Cited by 17 publications

References 34 publications

Improved Machine Learning Model for Urban Tunnel Settlement Prediction Using Sparse Data

Improved Machine Learning Model for Urban Tunnel Settlement Prediction Using Sparse Data

Built environment defect mapping, modeling, and management (D3M): A BIM-based integrated framework

Light-Trans YOLO: A Lightweight Network Based on Transformer for Defect Detection of Industrial Lace Surface

Contact Info

Product

Resources

About