Year-end Sale % OFF 
Abstract
This article introduces an automated data-labeling approach for generating crack ground truths (GTs) within concrete images. The main algorithm includes generating first-round GTs, pre-training a deep learning-based model, and generating second-round GTs. On the basis of the generated second-round GTs of the training data, a learning-based crack detection model can be trained in a self-supervised manner. The pre-trained deep learning-based model is effective for crack detection after it is re-trained using the second-round GTs. The main contribution of this study is the proposal of an automated GT generation process for training a crack detection model at the pixel level. Experimental results show that the second-round GTs are similar to manually marked labels. Accordingly, the cost of implementing learning-based methods is reduced significantly because data labeling by humans is not necessitated.
Highlights
Cracks appear on the surface of concrete structures owing to various causes, such as aging, environmental, and loading effects
Surface cracks are one of the earliest indicators of structural damage; crack monitoring has become a critical task in structural maintenance
The main contributions of this study are summarized below: 1. We introduce an algorithm that can perform automated data labeling for concrete images exhibiting cracks
Summary
Cracks appear on the surface of concrete structures owing to various causes, such as aging, environmental, and loading effects. Convolutional networks (FCNs) trained via end-to-end learning first demonstrated state-of-the-art performances in 2015 [16] They were investigated extensively to solve challenging problems in semantic segmentation, such as crack detection at the pixel level. On the basis of the well-known segmentation model SegNet [22], Zou et al [23] developed DeepCrack, which is a deep CNN, by learning high-level features for crack representation To obtain both sparse and continuous features in each scale, DeepCrack added skip layers to connect the encoder to the decoder. An algorithm for generating the GTs of concrete images that can be further used for training deep learning networks to perform crack detection is proposed . The main contribution of this study is the proposal of an automated labeling technique that involves a three-stage procedure, including first-round GT generation, pre-training of a U-Net-based model, and second-round GT generation.
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
