Crack Detection and Localization based on Spatio-Temporal Data using Residual Networks

Moreh, Fathalla; Lyu, Hao; Beth, Christian; Strohm, Steffen; Rizvi, Zarghaam Haider; Wuttke, Frank; Renz, Matthias

doi:10.1145/3538712.3538743

Cited by 2 publications

(5 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The following work showed that an advanced network, i.e. ResNet, facilitated better performance and reduced the required computational resource on the same dataset used for training the model with 1D-VGG-net 14 . ResNets are characterised by their skip connections to enable identity mapping and residual learning 52 .…”

Section: The Encoder Networkmentioning

confidence: 96%

“…Each sample contains a randomly assigned crack (or no crack). The previous studies built a small dataset and illustrated the influence of crack size on the model's performance 13,14 . Specifically, much smaller cracks are more difficult to detect than larger cracks.…”

Section: The Crack Detection Datasetmentioning

confidence: 99%

“…In this work, the proposed models follow the encoder-decoder architecture 13,14 . Different backbone networks were tested for the encoder.…”

Section: Deep Neural Network For Crack Detectionmentioning

confidence: 99%

See 2 more Smart Citations

Deep neural networks for crack detection inside structures

Moreh,

Lyu,

Rizvi

et al. 2024

Sci Rep

Self Cite

View full text Add to dashboard Cite

Crack detection is a long-standing topic in structural health monitoring. Conventional damage detection techniques rely on intensive, time-consuming, resource-intensive intervention. The current trend of crack detection emphasizes using deep neural networks to build an automated pipeline from measured signals to damaged areas. This work focuses on the seismic-wave-based technique of crack detection for plate structures. Previous work proposed an encoder–decoder network to extract crack-related wave patterns from measured wave signals and predict crack existence on the plate. We extend previous work with extensive experiments on different network components and a data preprocessing strategy. The proposed methods are tested on an expanded crack detection dataset. We found that a robust backbone network, such as Densely Connected Convolutional Network (DenseNet) can effectively extract the features characterizing cracks of wave signals, and by using the reference wave field for normalization, the accuracy of detecting small cracks can be further improved.

show abstract

Section: The Encoder Networkmentioning

confidence: 96%

Section: The Crack Detection Datasetmentioning

confidence: 99%

See 1 more Smart Citation

Deep neural networks for crack detection inside structures

Moreh,

Lyu,

Rizvi

et al. 2024

Sci Rep

Self Cite

View full text Add to dashboard Cite

show abstract

“…The following work showed that an advanced network, i.e. ResNet, facilitated better performance and reduced the required computational resource on the same dataset used for training the model with 1D-VGG-net 58 . ResNets are characterised by their skip connections to enable identity mapping and residual learning 50 .…”

Section: The Encoder Networkmentioning

confidence: 96%

“…The architecture of a part of these models was introduced in 56,58 . These models were also further modified and extended in this work.…”

Section: Deep Neural Network For Crack Detectionmentioning

confidence: 99%

Deep Neural Networks for Crack Detection Inside Structures

Moreh,

Lyu,

Rizvi

et al. 2023

Preprint

Self Cite

View full text Add to dashboard Cite

Crack detection is a long-standing topic in structural health monitoring. Conventional damage detection techniques rely on intensive, time-consuming, resource-intensive intervention. The current trend of crack detection emphasizes using deep neural networks to build an automated pipeline from measured signals to damaged areas. This work focuses on the seismic-wave-based technique of crack detection for plate structures. Previous work proposed an autoencoder network to extract crack-related wave patterns from measured wave signals and predict crack existence on the plate. We extend previous work with extensive experiments on different network components and a data preprocessing strategy. The proposed methods are tested on an expanded crack detection dataset. We found that a robust backbone network, such as DenseNet can effectively extract the features characterizing cracks of wave signals, and by using the reference wave field for normalization, the accuracy of detecting small cracks can be further improved.

show abstract

Crack Detection and Localization based on Spatio-Temporal Data using Residual Networks

Cited by 2 publications

References 10 publications

Deep neural networks for crack detection inside structures

Deep neural networks for crack detection inside structures

Deep Neural Networks for Crack Detection Inside Structures

Contact Info

Product

Resources

About