A novel approach to assess the seismic performance of deteriorated bridge structures by employing UAV‐based damage detection

Traditional manual inspection technology has the problems of high risk, low efficiency, and being time-consuming in bridge safety management. The unmanned aerial vehicle (UAV)-based detection technology is widely used in bridge structure safety monitoring. However, the existing deep learning-based concrete crack identification method has great limitations in dealing with complex background and tiny cracks in bridge structures. To address these problems, this study designs a crack pixel-level high-performance segmentation model for bridge concrete cracks that is suitable for UAV detection scenarios using machine vision (MV) and deep learning (DL) algorithms. First, considering the high requirements for the computing performance of the MV-based model for UAV-based detection, the ResNet-18-based lightweight convolutional neural network is used to represent the traditional large-scale backbone network of the pyramid scene parsing network (PSPNet) to develop a high-performance crack automatic identification model. Then, considering that bridge concrete cracks have the characteristics of subtle shapes and complex backgrounds, the spatial position self-attention module is inserted into the PSPNet to improve its detection accuracy. A concrete bridge is used for the case study, and a dataset of cracks in bridge concrete structures collected by UAVs is constructed and used for model training. The experimental results show that the loss function of the developed method in the training process results in a smooth decline, and the developed algorithm achieves the evaluation indicators of 0.9008 precision, 0.8750 recall, 0.8820 accuracy, and 0.9012 IOU on the bridge concrete crack dataset, which are significantly higher than other state-of-the-art baseline methods. In addition, four common UAV bridge detection scenarios, including low light, complex crack forms, high background roughness, and complex background scenes, are used to further test the crack detection ability of the developed crack identification model. The experimental results show that the proposed crack identification method can effectively overcome interference and real-size pixel-level segmentation of crack morphology. In addition, it also achieved a detection efficiency of 35.04 FPS, which shows that the real-time detection ability of the method has good applicability in the UAV detection scene.

Section: Introductionmentioning

confidence: 99%

Pixel‐Level Crack Identification for Bridge Concrete Structures Using Unmanned Aerial Vehicle Photography and Deep Learning

Song,

Liu,

Yuan

2024

“…When the application of the adhesives is related to structural parts demanding in terms of strength and/or fatigue requirements, the concern of the structural monitoring of the bonding integrity can rise [1][2][3][4]. Structural health monitoring (SHM) has received growing interest with multiple applications oriented to the development of damage/defect detection systems able to allow the transition from scheduled maintenance to condition-based maintenance in several areas [5][6][7][8]. An efective SHM method is the result of a reliable mathematical description of the structural part under control and the subsequent implementation of this model into a monitoring technique [9][10][11].…”

Section: Introductionmentioning

confidence: 99%

Two Dimensional Finite Difference Model with a Singularity Attenuation Factor for Structural Health Monitoring of Single Lap Joints

Nicassio

Vergallo

Vitolo

et al. 2023

A finite difference algorithm that evaluates the health conditions of a bonded joint is presented and discussed. The mathematical formulation of the problem is developed, paying particular attention to the singularity around the corners of the joint and implementing an original discretisation method of the partial differential equations governing the propagation of the elastic waves. The equations are solved under the only hypothesis of a bidimensional field. The algorithm is sensible to defects in the bonded joint and can be used as an effective structural health monitoring tool, as proven by the experiments that show close agreement with the numerical simulations.

“…Terefore, UAVs play an important role in the entire life cycle of engineering structures. Te daily maintenance of various structures, such as dams [15], bridges [16], roads [17], and buildings [18], can be realized by UAVs, which collect visual data from inaccessible regions [19]. In addition, UAVs can accomplish faster structural damage inspection and assessment in disaster areas because of the inconvenience of trafc after an emergency [20].…”

Section: Introductionmentioning

confidence: 99%

Feasibility of Accurate Point Cloud Model Reconstruction for Earthquake-Damaged Structures Using UAV-Based Photogrammetry

Shan

Zhu

2023

Camera-enabled unmanned aerial vehicles (UAVs) provide a promising technique to considerably speed up the inspection and visual data collection from regions that may otherwise be inaccessible. In addition, the technology of image-based 3D reconstruction can generate a point cloud model using images captured by UAVs. However, the performance of the point cloud modeling may be affected by multiple factors, such as the modeling software, ground control points (GCPs), and UAV flight modes. In this study, three common software packages were compared, and Pix4Dmapper was considered a suitable software for point cloud modeling for earthquake-damaged buildings. The accuracy and resolution of point cloud models are usually evaluated by root mean square error (RMSE) and ground sampling distance (GSD). The effects of the main factors, including the number of GCPs, distribution of GCPs, flight manner of the UAV, and distance from the UAV to the target, were investigated on the basis of two real-world multistory earthquake-damaged structures. The influence rules of the main factors revealed that a close range, automatic flight mode of the UAV, a large number of GCPs, and a relatively wide distribution of the GCPs may generate a point cloud model with low computational costs, high accuracy, and high resolution. In the particular illustration example here, the RMSE is 6.78 mm while the GSD is 1.60 mm. Finally, rapid structural damage inspection was demonstrated using an accurate point cloud model and compared with the inspection results of a total station and terrestrial laser scanner point cloud models. The comparison of different inspection results showed that the relative errors were relatively acceptable within 4%.