Pixel‐level multicategory detection of visible seismic damage of reinforced concrete components

Abstract

The detection of visible damage (i.e., cracking, concrete spalling and crushing, reinforcement exposure, buckling and fracture) plays a key role in postearthquake safety assessment of reinforced concrete (RC) building structures. In this study, a novel approach based on computer-vision techniques was developed for pixel-level multicategory detection of visible seismic damage of RC components. A semantic segmentation database was constructed from test photos of RC structural components. Series of datasets were generated from the constructed database by applying image transformations and data-balancing techniques at the sample and pixel levels. A deep convolutional network architecture was designed for pixel-level detection of visible damage. Two sets of parameters were optimized separately, one to detect cracks and the other to detect all other types of damage. A postprocessing technique for crack detection was developed to refine crack boundaries, and thus improve the accuracy of crack characterization. Finally, the proposed vision-based approach was applied to test photos of a beam-to-wall joint specimen. The results demonstrate the accuracy of the vision-based approach to detect damage, and its high potential to estimate seismic damage states of RC components.

Multicategory damage detection and safety assessment of post‐earthquake reinforced concrete structures using deep learning

“…For example, in the specification (AQSIQ, 2009), the extent of damage is defined as follow: (a) the visible cracks on the surface of the floated coat or mortar joint; (b) crack width less than 0.5 mm; (c) crack width more than 0.5 mm and surface spalling; (d) the surface of components severe spalling and rebars exposed obviously; and (e) parts of the components crushed and rebars buckled. According to the damage description in literature (Miao et al., 2021) and the simulated damage contour plots under different damage levels, the corresponding relationship between damage level (none damage, minor damage, moderate damage, severe damage, and collapse) and the most severe damage (fine crack, wide crack, concrete spalling, exposed rebar, and buckled rebar) on the component surface was established as shown in Figure 4.…”

“…Although many studies have great success in damage detection (Ghosh Mondal et al., 2020) and segmentation (S. Li et al., 2019; Yang et al., 2018), the research of how to correlate the detection (segmentation) results with damage states of components (Cheng et al., 2021) and how to correlate damage states of components with structural damage states effectively is relatively lacking (Gulgec et al., 2019; Miao et al., 2021).…”

Multistage semisupervised active learning framework for crack identification, segmentation, and measurement of bridges

“…Nowadays, machine learning (ML) and deep learning (DL) have been at the forefront of a plethora of research activities in many disciplines, where topics in structural engineering, for example, system parameter identification (Perez‐Ramirez et al., 2016), bolt loosening detection (Yang et al., n.d.), and crack detection, significantly benefit from these technologies. In order to reduce the cost and safety risk while improving detection accuracy and automation, researchers have proposed a variety of image detection technologies combining ML/DL with CV to recognize cracks (Kong et al., 2021; Liang, 2019; Miao et al., 2021; Żarski et al., 2021; Zhang & Yuen, 2021). In general, vision‐based crack detection has three main tasks: identification, localization, and segmentation.…”

A deep learning‐based image captioning method to automatically generate comprehensive explanations of bridge damage

“…Notable studies include research on the detection of cracks in concrete Deng et al, 2020;Dung, 2019;Jang et al, 2021;Jiang & Zhang, 2020;Kong et al, 2021;Lee et al, 2019;Y. Liu et al, 2019;Miao et al, 2021;Sajedi & Liang, 2021; and asphalt pavements (F. Guo et al, 2021;Li et al, 2020;Maeda et al, 2018;Maeda et al, 2021) and in the detection of defects and leaks in tunnels (Huang et al, 2018;Xue & Li, 2018). Further, research has been conducted on evaluating the soundness of structures using vibration (Avci et al, 2021;Gutierrez Soto & Adeli, 2019;, 2018aShrestha & Dang, 2020) and on assessing the internal damage to concrete using nondestructive testing (Chun & Hayashi 2021;Chun, Ujike, et al, 2020;Erdal et al, 2018;Sirca et al, 2018).…”