Abstract: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.

Automated crack segmentation in close-range building façade inspection images using deep learning techniques

Two-Stream Boundary-Aware Neural Network for Concrete Crack Segmentation and Quantification

Ultrasonic Array Tomography-Oriented Subsurface Crack Recognition and Cross-Section Image Reconstruction of Reinforced Concrete Structure Using Deep Neural Networks

A Three-Step Computer Vision-Based Framework for Concrete Crack Detection and Dimensions Identification

Automatic detection of building surface cracks using UAV and deep learning‐combined approach

Inspecting Buildings Using Drones and Computer Vision: A Machine Learning Approach to Detect Cracks and Damages

A Deep Learning Approach for Surface Crack Classification and Segmentation in Unmanned Aerial Vehicle Assisted Infrastructure Inspections

UAV-Driven Structural Crack Detection and Location Determination Using Convolutional Neural Networks

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

Structural Crack Detection Using Deep Learning–based Fully Convolutional Networks

Real‐time crack assessment using deep neural networks with wall‐climbing unmanned aerial system

Vision and Deep Learning-Based Algorithms to Detect and Quantify Cracks on Concrete Surfaces from UAV Videos

UAV vision detection method for crane surface cracks based on Faster R-CNN and image segmentation

Deep-Learning- and Unmanned Aerial Vehicle-Based Structural Crack Detection in Concrete

Transfer Learning-Based Crack Detection by Autonomous UAVs

Application of Deep Learning and Unmanned Aerial Vehicle on Building Maintenance

Autonomous Inspection of High-Rise Buildings for Façade Detection and 3D Modeling Using UAVs

Hierarchical Convolutional Neural Network with Feature Preservation and Autotuned Thresholding for Crack Detection

An improved deep learning convolutional neural network for crack detection based on UAV images

Learning-Based Defect Recognitions for Autonomous UAV Inspections