Figure - available via license: Creative Commons Attribution 4.0 International
Content may be subject to copyright.
Source publication
Automatic crack detection from images is an important task that is adopted to ensure road safety and durability for Portland cement concrete (PCC) and asphalt concrete (AC) pavement. Pavement failure depends on a number of causes including water intrusion, stress from heavy loads, and all the climate effects. Generally, cracks are the first distres...
Contexts in source publication
Context 1
... groups of {1, 2, 3, 4}, {1, 2, 4, 8}, {2, 4, 8, 16} are tested based on public database CFD and AigleRN. As shown from the experimental results in Tables 2 and 3, group of {2, 4, 8, 16} can obtain the highest accuracy on both databases. The reason is that a large dilation rate can get more context information of the cracks for the relatively wide or thin crack structure, which can improve the crack detection accuracy. ...Context 2
... dilation rate presented in Equation (1) Three groups of {1, 2, 3, 4}, {1, 2, 4, 8}, {2, 4, 8, 16} are tested based on public database CFD and AigleRN. As shown from the experimental results in Table 2 and Table 3, group of {2, 4, 8, 16} can obtain the highest accuracy on both databases. The reason is that a large dilation rate can get more context information of the cracks for the relatively wide or thin crack structure, which can improve the crack detection accuracy. ...Similar publications
Vehicle and driver detection in the highway scene has been a research hotspot in the field of object detection in recent years, and it is still a challenging problem in the research of traffic order and road safety. In this paper, we propose a novel end-to-end vehicle and driver detection method named VDDNet which is based on Cascade R-CNN and SENe...
Citations
... Recent work has facilitated automated classification, localization, and quantification of structural defects from image data (Cha et al., 2017;Chen and Jahanshahi, 2017;Cha et al., 2018;Attard et al., 2019;Liu et al., 2019;Li et al., 2020Li et al., , 2024a. DL-based techniques have shown promise in detecting cracks in buildings (Perez et al., 2019;Jiang et al., 2021), bridges (Dais et al., 2021;Hallee et al., 2021;Loverdos and Sarhosis, 2022), tunnels (Liao et al., 2022a;Protopapadakis et al., 2019), and roads (Fan et al., 2020). For example, recent studies reported the successful detection of cracks with widths greater than ≤ 1 mm (Liao et al., 2022b;Mohammadi et al., 2019). ...
Ageing structures require periodic inspections to identify structural defects.
Previous work has used geometric distortions to locate cracks in synthetic
masonry bridge point clouds but has struggled to detect small cracks. To
address this limitation, this study proposes a novel 3D multimodal feature,
3DMulti-FPFHI, which combines a customized Fast Point Feature Histogram
(FPFH) with an intensity feature. This feature is integrated into the PatchCore anomaly detection algorithm and evaluated through statistical and
parametric analyses. The method is further evaluated using point clouds of
a real masonry arch bridge and a full-scale experimental model of a concrete
tunnel. Results show that the 3D intensity feature enhances inspection quality by improving crack detection; it also enables the identification of water
ingress which introduces intensity anomalies. The 3DMulti-FPFHI outperforms FPFH and a state-of-the-art multimodal anomaly detection method.
The potential of the method to address diverse infrastructure anomaly detection scenarios is highlighted by the minimal requirements for data compared
to learning-based methods. The code and related point cloud dataset are
available at https://github.com/Jingyixiong/3D-Multi-FPFHI.
... Recent work has facilitated automated classification, localization, and quantification of structural defects from image data (Cha et al., 2017;Chen and Jahanshahi, 2017;Cha et al., 2018;Attard et al., 2019;Liu et al., 2019;Li et al., 2020Li et al., , 2024a. DL-based techniques have shown promise in detecting cracks in buildings (Perez et al., 2019;Jiang et al., 2021), bridges (Dais et al., 2021;Hallee et al., 2021;Loverdos and Sarhosis, 2022), tunnels (Liao et al., 2022a;Protopapadakis et al., 2019), and roads (Fan et al., 2020). For example, recent studies reported the successful detection of cracks with widths greater than ≤ 1 mm (Liao et al., 2022b;Mohammadi et al., 2019). ...
Ageing structures require periodic inspections to identify structural defects. Previous work has used geometric distortions to locate cracks in synthetic masonry bridge point clouds but has struggled to detect small cracks. To address this limitation, this study proposes a novel 3D multimodal feature, 3DMulti-FPFHI, that combines a customized Fast Point Feature Histogram (FPFH) with an intensity feature. This feature is integrated into the PatchCore anomaly detection algorithm and evaluated through statistical and parametric analyses. The method is further evaluated using point clouds of a real masonry arch bridge and a full-scale experimental model of a concrete tunnel. Results show that the 3D intensity feature enhances inspection quality by improving crack detection; it also enables the identification of water ingress which introduces intensity anomalies. The 3DMulti-FPFHI outperforms FPFH and a state-of-the-art multimodal anomaly detection method. The potential of the method to address diverse infrastructure anomaly detection scenarios is highlighted by the minimal requirements for data compared to learning-based methods. The code and related point cloud dataset are available at https://github.com/Jingyixiong/3D-Multi-FPFHI.
... Recently, ED-CNNs have been developed for semantic image segmentation [18]. Motivated by these accomplishments, numerous recent investigations have devised ED-CNN-based models aimed at automatic semantic segmentation of concrete cracks [19,20]. ...
The longevity and safety of concrete precast crane beams significantly impact the operational integrity of industrial infrastructure. Assessment of surface cracks development in concrete structural elements during laboratory tests is performed mainly by applying standard tools such as linear-variable-differential transformers and strain gauges. This paper presents a novel assessment methodology combining deep convolutional neural network for image segmentation with digital image correlation method to evaluate the structural health of precast crane beams after more than fifty years of service. The study first outlines the adaptation of the deep learning U-Net architecture for detecting and segmentation of surface cracks in crane beams. Concurrently, DIC technique is employed to measure surface strains and displacements under load. The integration of these technologies enables a non-destructive, accurate, and detailed analysis, facilitating early detection of deterioration that may compromise structural safety. Initial results from field tests validate the effectiveness of our approach, demonstrating its potential as a tool for predictive maintenance of aging industrial infrastructure.
... And inserted a spatial focus to reuse features. 9. UHDN [31]: UHDN propose an encoder-decoder architecture with hierarchical feature learning and dilated convolution and design hierarchical feature learning module. 10. ...
Roads frequently experience cracks. It adversely impact the safe passage of vehicles and pedestrians, and have the potential to alter the road’s structure. To address this issue, we propose a novel crack detection network. The network constructs multi-channel attention and enhanced information interaction mechanisms to capture more granular semantic information. In our network, each convolutional layer is followed by a convolution combining asymmetric convolutions and criss-cross attention to enhance the feature maps post-convolution. This is followed by spatial and channel reconstruction convolutions and shuffle attention to optimize the generated side-output features. By extensively mining features from the deep network and ingeniously integrating bottom-level and top-level features through a new feature fusion module. The network achieves precise crack prediction results. Extensive experiments on the general-purpose crack image datasets Crack500, CFD and DeepCrack demonstrate the model’s effectiveness. In these three datasets, F1-score values of 0.734, 0.635, and 0.881, MIoU values of 0.773, 0.726 and 0.888.
... In exploring alternative methods, Hoang and Nguyen [73] evaluated SVM, ANN, and RF and found that SVM performed best. Additionally, Fan et al. [124] proposed the U-Hierarchical Dilated Network (U-HDN), an end-to-end DL algorithm for crack detection. Tong et al. [125] integrated a fully convolutional network with a Gaussian-Conditional Random Field (G-CRF) for pavement defect detection. ...
Effective road pavement management is vital for maintaining the functionality and safety of transportation infrastructure. This review examines the integration of Machine Learning (ML) into Pavement Management Systems (PMS), presenting an analysis of state-of-the-art ML techniques, algorithms, and challenges for application in the field. We discuss the limitations of conventional PMS and explore how Artificial Intelligence (AI) algorithms can overcome these shortcomings by improving the accuracy of pavement condition assessments, enhancing performance prediction, and optimizing maintenance and rehabilitation decisions. Our findings indicate that ML significantly advances PMS capabilities by refining data collection processes and improving decision-making, thereby addressing the intricacies of pavement deterioration. Additionally, we identify technical challenges such as ensuring data quality and enhancing model interpretability. This review also proposes directions for future research to overcome these hurdles and to help stakeholders develop more efficient and resilient road networks. The integration of ML not only promises substantial improvements in managing pavements but is also in line with the increasing demands for smarter infrastructure solutions.
... Mandal et al. utilized YOLO for crack detection in various datasets, including different types of cracks and defects [25]. Fan et al. [26] proposed a modified version of the UNet, incorporating dilated convolution modules and hierarchical feature learning modules to enhance the performance of crack detection. They introduced a method known as CrackGAN, drawing on the concept of generative adversarial networks, and utilized an asymmetric U-Net network as the backbone architecture to process images of arbitrary sizes. ...
Dams in their natural environment will gradually develop cracks and other forms of damage. If not detected and repaired in time, the structural strength of the dam may be reduced, and it may even collapse. Repairing cracks and defects in dams is very important to ensure their normal operation. Traditional detection methods rely on manual inspection, which consumes a lot of time and labor, while deep learning methods can greatly alleviate this problem. However, previous studies have often focused on how to better detect crack defects, with the corresponding image resolution not being particularly high. In this study, targeting the scenario of real-time detection by drones, we propose an automatic detection method for dam crack targets directly on high-resolution remote sensing images. First, for high-resolution remote sensing images, we designed a sliding window processing method and proposed corresponding methods to eliminate redundant detection frames. Then, we introduced a Gaussian distribution in the loss function to calculate the similarity of predicted frames and incorporated a self-attention mechanism in the spatial pooling module to further enhance the detection performance of crack targets at various scales. Finally, we proposed a pruning-after-distillation scheme, using the compressed model as the student and the pre-compression model as the teacher and proposed a joint distillation method that allows more efficient distillation under this compression relationship between teacher and student models. Ultimately, a high-performance target detection model can be deployed in a more lightweight form for field operations such as UAV patrols. Experimental results show that our method achieves an mAP of 80.4%, with a parameter count of only 0.725 M, providing strong support for future tasks such as UAV field inspections.
... Therefore, segmentation-based crack detection [11,[14][15][16][17][18] has received more and more attention. Fan et al [19] proposed an improved version of UNet for crack detection by adding a multiple dilation module and a hierarchical feature learning module to the original architecture. Zou et al [20] proposed a novel encoder-decoder method based on the structure of SegNet, which uses a hierarchical convolution stage to learn multi-scale crack features. ...
Crack detection is an important task to ensure structural safety. Traditional manual detection is extremely time-consuming and labor-intensive. However, existing deep learning-based methods also commonly suffer from low inference speed and continuous crack interruption. To solve the above problems, a novel bilateral crack detection network (BiCrack) is proposed for real-time crack detection tasks. Specifically, the network fuses two feature branches to achieve the best trade-off between accuracy and speed. A detail branch with a shallow convolutional layer is first designed. It preserves crack detail to the maximum and generates high-resolution features. Meanwhile, the semantic branch with fast-downsampling strategy is used to obtain enough high-level semantic information. Then, a simple pyramid pooling module (SPPM) is proposed to aggregate multi-scale context information with low computational cost. In addition, to enhance feature representation, an attention-based feature fusion module (FFM) is introduced, which uses space and channel attention to generate weights, and then fuses input fusion features with weights. To demonstrate the effectiveness of the proposed method, it was evaluated on 5 challenging datasets and compared with state-of-the-art crack detection methods. Extensive experiments show that BiCrack achieves the best performance in the crack detection task compared to other methods.
... This approach allows the model to capture detailed features without the limitation of fully connected layers.The UNet architecture [43] further improves segmentation accuracy through its innovative use of skip connections and upsampling techniques, resulting in a more detailed and accurate representation of crack features. Extensions to UNet, as seen in models like Deepcrack [44] and UHDN [45], introduce novel methods for feature fusion at various levels. ...
Deep-learning-based crack identification has emerged as a prominent research area in structural health monitoring. Although the detection of common cracks has been the predominant focus in previous studies, the identification of tiny cracks has often been neglected. Efficiently managing thin cracks is vital, because they can threaten the overall structural integrity over time if left unaddressed. We address this gap by targeting thin cracks within a broad category of crack types. We introduce a fine-crack-detection algorithm that efficiently detects both common and tiny cracks. Owing to the limited availability of publicly accessible datasets specifically focused on thin cracks, we collect images of fine cracks to train and evaluate our algorithm. To validate the efficiency of our method, we conduct experiments on three publicly available crack datasets and our private dataset. Compared with the baseline neural network, our proposed approach demonstrates superior performance across all evaluation metrics. Furthermore, our model exhibits impressive generalization ability across the datasets, with the F1 score and mean intersection over union improving by 22.42% and 28.07%, respectively. Notably, our observations indicate that the advantages of the proposed method become more pronounced as the dataset size increases.
... Sun et al [18] proposed the DMA model in 2022, which incorporates dilated convolutions in the encoder stage, introduces the APSS module to capture global pooling, and finally integrates attention mechanisms in the decoder stage [24]. Fan et al designed an automated network that creates multiple dilation blocks in the encoding and decoding stages to conduct accurate evaluations of highway leakage cracks at high speed [25]. Hong et al conducted research on UAV remote sensing highway pavement crack images based on the U-Net model. ...
Efficient and precise identification of road pavement cracks contributes to better evaluation of road conditions. In practical road maintenance and safety assessment, traditional manual crack detection methods are time-consuming, physically demanding, and highly subjective. In addition, crack recognition based on image processing techniques lacks robustness. In this paper, a multi-branch feature fusion road crack segmentation network model (DTPC) based on deep convolution and transformer modules is proposed. The model is used for pixel-level segmentation of road crack images, which is a good solution to the existing needs and helps to repair dangerous cracks promptly in the follow-up work to prevent serious disasters due to crack breakage. Firstly, combine deep convolution with transformer modules to achieve precise local extraction and global contextual feature extraction. Secondly, a dual-channel attention mechanism is employed to help the model better address information loss and positional offset issues. Finally, three-branch outputs are fused to obtain prediction maps that intuitively determine recognition results. The proposed model is tested for accuracy using a dedicated road pavement crack dataset. Results show that compared to mainstream models such as SegFormer, HRNet, PSPNet, and fully convolutional network, the DTPC model achieves the highest MIoU score (86.72%) and F1 score (92.49%).
... To enhance the performance of UNet, Liu et al. [37] proposed a two-step CNN-based strategy, initially utilizing the YOLOv3 algorithm for crack location, succeeded by UNet for detailed segmentation. Inspired by UNet, Fan et al. [38] proposed a U-HDN network that integrates multiscale features for crack detection, and Zhang et al. [39] proposed an improved pixel-level surface crack detection model called CrackUNet. Furthermore, incorporating UNet within Generative Adversarial Networks (GAN) has shown promising results in pavement crack detection [40]. ...
Surface fractures are a significant problem in engineering structures like buildings and roads. Therefore, detecting such cracks is essential for assessing damage and maintaining these structures. The emergence of deep learning techniques has significantly enhanced the capability to detect surface cracks. Convolutional Neural Networks (CNNs) are predominantly used for this task, but recently introduced transformer architectures could offer improvements. In this research, we developed software that integrates nine advanced models and various activation functions to evaluate their effectiveness in detecting pavement cracks. The evaluation is based on the models’ accuracy, complexity, and stability. We generated 711 images, each
pixels, with crack labels, selected the most effective loss function, and compared the performance metrics of both validation and test datasets. We also examined the data details and evaluated the segmentation results for each model. Our results show that transformer-based models are more likely to converge during training and achieve higher accuracy, albeit with increased memory usage and reduced processing speed. Considering both accuracy and efficiency, SwinUNet outperforms the other two transformers and emerges as the superior choice among the nine evaluated models. We further confirm our conclusions with two public crack datasets. These findings shed light on the capabilities of various deep-learning models for surface crack detection and offer guidance for future applications in the field.