检索结果-内蒙古大学图书馆

Lightweight multi-scale encoder- decoder network with locally enhanced attention mechanism for concrete crack segmentation

引用

MEASUREMENT SCIENCE AND TECHNOLOGY 2025年第2期36卷 025021-025021页

作者： Dong, Shuai Cao, Jin Wang, Yunlong Ma, Jia Kuang, Zhiwei Zhang, Zhenhao Changsha Univ Sci & Technol Sch Civil Engn Changsha 410114 Peoples R China Saarland Univ Inst Appl Mech D-66123 Saarbrucken Germany

concrete surface crack detection and maintenance are crucial for ensuring structural safety. Deep learning-based techniques for detecting concrete cracks have become popular due to the quick advancement of artificial intelligence. However, the actual uses of these methods are limited due to issues like large model sizes and significant dependence on powerful computing hardware. To address these issues, this paper presents a lightweight multi-scale encoder-decoder network (LMED-Net) for crack detection of concrete structures. LMED-Net employs MobileNetV2 as the encoder for the initial feature extraction. A multi-scale feature extraction (MFE) module is developed and serially attached after the encoder for refining feature extraction. Finally, to strengthen the network's perception of pixels surrounding the cracks, a novel enhanced attention mechanism (EAM) is deployed in the decoder. By improving the network's attention to information within the crack regions, this mechanism keeps contextual information from being lost. Comparative experimental results show that the proposed network achieves an F1 score (F1) of 60.32% and a mean intersection over union (mIoU) of 71.04% on the crack forest dataset. On the Deepcrack dataset, the F1 and mIoU increase to 79.09% and 81.85% respectively. Notably, LMED-Net performs exceptionally well in crack segmentation since its model size and parameters count are much smaller than those of other image segmentation methods. Furthermore, ablation studies further validate the effectiveness of the proposed MFE module and EAM.

关键词： concrete crack detection semantic segmentation lightweight network encoder-decoder network local enhanced attention mechanism

来源：评论

学校读者我要写书评

暂无评论

Automated vision-based concrete crack measurement system

引用

MEASUREMENT 2025年第PartA期242卷

作者： Giulietti, Nicola Revel, Gian Marco Chiariotti, Paolo Univ Pavia Dipartimento Ingn Ind & Informaz Via Adolfo Ferrata 5 I-27100 Pavia Italy Univ Politecn Marche Dept Ind Engn & Math Sci Via Brecce Bianche 12 I-60131 Ancona Italy Politecn Milan Dept Mech Engn Via Masa 1 I-20156 Milan Italy

Maintaining and ensuring the safety of ageing infrastructures is becoming of uttermost importance. This is particularly critical in Europe, where several infrastructures like tunnels and bridges are approaching their end-of-life. Such a condition justifies the need for a prompt recognition of the deterioration status of these structures as well as the identification of strategies for effective inspection and monitoring operations. This paper focuses on cracks, and proposes a vision-based measurement system aimed at precisely and accurately detecting cracks on concrete surfaces and measuring their geometric features. The system mainly leverages RGB-D cameras to make the pixel to real world measurement units conversion possible (by assessing the camera to target relative distance and pose), and exploits an image processing pipeline consisting of the following steps: (a) an artificial intelligence-based semantic segmentation model for identifying cracks on the target structure;(b) a ridge detection algorithm for identifying the centre line of the cracks;(c) a geometric feature extraction algorithm to extract key metrics of the crack, like its width and length. Variance analysis of the crack mean width measurement shows a repeatability standard deviation of 0.01 mm and a reproducibility standard deviation of 0.02 mm, with a Type A expanded uncertainty of 0.004 mm. The system performance is discussed in the paper referring both to lab testing activities, targeted to the assessment of its metrological characteristics, and to a real field application, i.e. concrete inspection in a tunnel. This latter application is of particular interests as it is targeted to demonstrate the robustness of the approach in harsh environments (i.e., strong presence of dust, total absence of natural light) and on surfaces whose roughness generates important light gradients and shadows (e.g., sprayed concrete).

关键词： Vision-based measurement system concrete crack measurement concrete crack detection Surface crack geometric feature extraction Uncertainty analysis

来源：评论

学校读者我要写书评

暂无评论

Real-time detection and geometric analysis algorithm for concrete cracks based on the improved U-net model

引用

JOURNAL OF REAL-TIME IMAGE PROCESSING 2024年第4期21卷 139-139页

作者： Zhang, Qian Zhang, Fan Liu, Hongbo Wang, Longxuan Chen, Zhihua Guo, Liulu China Construct Sixth Engn Bur Corp Tianjin 300072 Peoples R China Tianjin Univ State Key Lab Hydraul Engn Simulat & Safety Tianjin 300072 Peoples R China Hebei Univ Engn Dept Civil Engn Handan 056000 Peoples R China

Aiming at complex operation problems, low precision and poor robustness of traditional concrete crack detection methods, a real-time concrete crack detection and geometric analysis algorithm based on the improved U-net model is proposed. First, the efficient channel attention (ECA) module is embedded in the U-net model to reduce the loss of target information. The DenseNet network is used instead of the VGG16 network in the U-net basic model architecture, making transmitting features and gradients more effective. Then, based on the improved U-net model, the concrete crack detection experiment is performed. The experimental results indicate that the improved U-net model has 91.56% pixel accuracy (PA), 80.12% mean intersection over union (mIoU), 84.89% recall and 88.10% F1_score. The mIoU, PA, recall and F1_score of the improved U-net model increased by 17.39%, 7.82%, 2.62% and 5.10%, respectively, compared with the original model. Next, the real-time detection experiment of concrete cracks is performed based on the improved U-net model. The FPS of the improved model is the same as that of the original model and reaches 42. Finally, the geometric analysis of concrete cracks is performed based on the detection results of the improved U-net model. The area, density, length and average width information of concrete cracks are effectively extracted. The research results indicate that the detection effect of this study's model on concrete cracks is considerably improved and that the model has good robustness. The model proposed in this study can achieve intelligent real-time and accurate identification of concrete cracks, which has broad application prospects.

关键词： concrete crack detection Improved U-net model Real-time detection Geometric analysis Deep learning Computer vision

来源：评论

学校读者我要写书评

暂无评论

Classification and detection of Various Structural cracks Using Deep Learning Approach 10th

Classification and Detection of Various Structural Cracks Us...

引用

10th European Workshop on Structural Health Monitoring (EWSHM)

作者： Vundekode, Narasimha Reddy Kalapatapu, Prafulla Pasupuleti, Venkata Dilip Kumar Mahindra Univ Ecole Cent Coll Engn Hyderabad India

ISBN: (纸本)9783031072581;9783031072574

concrete has evolved to be the most used construction material all over the world from minor to major structures. The accelerated deterioration of the structure can be reduced if the cracks are detected at their early stage, and it will be more effective if the crack type is also known. Development of crack and its propagation if unnoticed overtime can lead to minor to major damage. The proposed study targets to revolutionize the visual inspection process using Deep Learning (DL) algorithms. The most important part of training a DL algorithm is the data, hence the major challenge for the study is collection of various images consisting different types of cracks from the visual inspections carried manually during real-time projects. Overall, 3500 images were collected and out of which considering the computational complexity the proposed model was developed based on a 500-image sub dataset. The pre-trained DL model has demonstrated the desirable accuracies while no pre-trained model has dropped the accuracy. In practice, the paper also focuses on the current achievements and limitations of the existing methods and as well as opens lot of avenues for the upcoming researchers in the cross disciplinary fields of Structural and Computer Science.

关键词： concrete crack detection crack type Deep learning algorithm Computer vision Structural health monitoring

来源：评论

学校读者我要写书评

暂无评论

Real-Time Surface Damage Analysis of concrete Structures using Python-based Image Analysis 17

Real-Time Surface Damage Analysis of Concrete Structures usi...

引用

Conference on Nondestructive Characterization and Monitoring of Advanced Materials, Aerospace, Civil Infrastructure, and Transportation XVII

作者： Jha, Shivam Kaur, Amardeep Southern Illinois Univ Edwardsville Edwardsville IL 62026 USA

ISBN: (纸本)9781510660816;9781510660823

Aging infrastructure worldwide has led to an interest in using innovative solutions for non-destructive damage assessment. The traditional in-person assessment of large-scale infrastructure like dams and bridges etc. can be very time and resource-consuming. With advances in technology, image processing and machine learning have shown promise in providing alternate ways of damage assessment in large-scale and difficult-to-access infrastructure. However, this approach has mostly been applied passively and not in real-time. The work presented here describes a supervised machine learning-based approach to damage analysis of concrete structures. Python programming language is used to write and train algorithms to provide a real-time damage analysis. In addition to crack detection, the dimensional analysis provides additional information regarding an existing or developing crack. With this type of real-time information, timely action can be taken regarding decisions like performing repairs or decommissioning a structure for public safety.

关键词： Image Analysis concrete crack detection Structural Health Monitoring Machine Learning Python programming

来源：评论

学校读者我要写书评

暂无评论

crack depth measurement and key points of accurate identification in concrete structures: a review

引用

NONDESTRUCTIVE TESTING AND EVALUATION 2024年第8期39卷 2073-2107页

作者： Wang, Yulei Jia, Yu Liang, Jiahui Tang, Lei Guan, Fuhai Li, Po Zhang, Shengxing Tian, Hui Nanjing Hydraul Res Inst Mat & Struct Engn Dept Nanjing Peoples R China Natl Key Lab Water Disaster Prevent Nanjing Peoples R China

The cracks with a width of more than 0.3mm can seriously affect the safety and stability of concrete structures. Presently, there are many techniques for measuring the crack depth, each with its own advantages and disadvantages. However, due to the technical limitations and different applicability of each method, their accuracy in actual crack detection is often insufficient. In this article, to improve the detection accuracy in technological applications, the mature concrete crack depth detection technologies are summarised. The currently popular non-destructive testing technologies are divided into two categories: stress wave detection technology and electromagnetic wave detection technology. The specific detection methods and development status of the two technologies are discussed. Further, based on the detection principles, the advantages and disadvantages of various detection methods are systematically compared. The application conditions of each detection method are explored, and their feasibility for different forms of cracks are comprehensively analysed. A set of optimal methods for crack depth detection is proposed. Besides, the key points for the accurate measurement of concrete crack depth are discussed.

关键词： concrete crack detection stress wave detection technology electromagnetic wave detection technology accurate identification points

来源：评论

学校读者我要写书评

暂无评论

Intelligent concrete Surface cracks detection using Computer Vision, Pattern Recognition, and Artificial Neural Networks

引用

Procedia Computer Science 2023年 217卷 52-61页

作者： Majid Mirbod Maryam Shoar

This paper presents an intelligent concrete surface crack detection model, using machine vision, pattern recognition, artificial neural networks, and histogram of oriented gradients to feature extraction. This is an applied research category that so far, has been used in face recognition systems, and scaled conjugate gradient to training purpose multilayer neural networks, to achieve the monitoring of concrete surfaces to increase the protection that these are the innovations of this research. The main purpose of this research is to provide a reliable method for inspecting sensitive structures such as bridges using a continuous intelligent inspection system consisting of a camera and intelligent software. For this purpose, we used the Utah State University image dataset that contains over 56,000 images of cracked and non-cracked concrete bridge decks, walls, and pavements. Moreover, it includes data on cracks as narrow as 0.06 mm and a width of 25 mm with a variety of obstructions, shadows, surface roughness, scaling, edges, holes, and background debris. The ANN proposed in this study can recognize cracking images from non-cracking images. The efficiency of this network is 84.88% and it reduces network error from 0.631 to 0.199 and gradient index from 25.5 to 0.0801. This model will be useful for structural health monitoring and other sensitive surfaces such as aircraft engine shells and train moving rails. Based on the obtained results, the difference between the proposed method and other neural networks such as Convolutional Neural networks is that the former is simpler in coding and needs cheaper hardware.

关键词： concrete crack detection Histogram of oriented gradients Pattern recognition Machine vision Scaled conjugate gradient Multilayer neural network

来源：评论

学校读者我要写书评

暂无评论

Automatic detection of cracks on concrete Surfaces in the Presence of Shadows

引用

SENSORS 2022年第10期22卷 3662-3662页

作者： Palevicius, Paulius Pal, Mayur Landauskas, Mantas Orinaite, Ugne Timofejeva, Inga Ragulskis, Minvydas Kaunas Univ Technol Dept Math Modelling Ctr Nonlinear Syst LT-51368 Kaunas Lithuania

Deep learning-based methods, especially convolutional neural networks, have been developed to automatically process the images of concrete surfaces for crack identification tasks. Although deep learning-based methods claim very high accuracy, they often ignore the complexity of the image collection process. Real-world images are often impacted by complex illumination conditions, shadows, the randomness of crack shapes and sizes, blemishes, and concrete spall. Published literature and available shadow databases are oriented towards images taken in laboratory conditions. In this paper, we explore the complexity of image classification for concrete crack detection in the presence of demanding illumination conditions. Challenges associated with the application of deep learning-based methods for detecting concrete cracks in the presence of shadows are elaborated on in this paper. Novel shadow augmentation techniques are developed to increase the accuracy of automatic detection of concrete cracks.

关键词： concrete crack detection deep learning convolution neural networks image classification image augmentation

来源：评论

学校读者我要写书评

暂无评论

Review of Non-Destructive Civil Infrastructure Evaluation for Bridges: State-of-the-Art Robotic Platforms, Sensors and Algorithms

引用

SENSORS 2020年第14期20卷 3954.-3954页

作者： Ahmed, Habib La, Hung Manh Gucunski, Nenad Univ Nevada Dept Comp Sci & Engn Adv Robot & Automat Lab Reno NV 89512 USA Rutgers State Univ Dept Civil & Environm Engn Piscataway NJ 08854 USA

The non-destructive evaluation (NDE) of civil infrastructure has been an active area of research in recent decades. The traditional inspection of civil infrastructure mostly relies on visual inspection using human inspectors. To facilitate this process, different sensors for data collection and techniques for data analyses have been used to effectively carry out this task in an automated fashion. This review-based study will examine some of the recent developments in the field of autonomous robotic platforms for NDE and the structural health monitoring (SHM) of bridges. Some of the salient features of this review-based study will be discussed in the light of the existing surveys and reviews that have been published in the recent past, which will enable the clarification regarding the novelty of the present review-based study. The review methodology will be discussed in sufficient depth, which will provide insights regarding some of the primary aspects of the review methodology followed by this review-based study. In order to provide an in-depth examination of the state-of-the-art, the current research will examine the three major research streams. The first stream relates to technological robotic platforms developed for NDE of bridges. The second stream of literature examines myriad sensors used for the development of robotic platforms for the NDE of bridges. The third stream of literature highlights different algorithms for the surface- and sub-surface-level analysis of bridges that have been developed by studies in the past. A number of challenges towards the development of robotic platforms have also been discussed.

关键词： non-destructive evaluation (NDE) structural health monitoring (SHM) electric resistivity (ER) sensors ground-penetrating radar (GPR) infrared (IR) thermography impact echo (IE) NDE sensor fusion convolutional neural network (CNNs) concrete crack detection rebar detection and localization

来源：评论

学校读者我要写书评

暂无评论

A New Approach for the detection of concrete cracks Based on Adaptive Morphological Filtering 6

A New Approach for the Detection of Concrete Cracks Based on...

引用

6th International Conference on Fuzzy Systems and Data Mining (FSDM)

作者： Hou, Huan Lin, Weiguo Beijing Univ Chem Technol Coll Informat Sci & Technol Beijing 100029 Peoples R China

ISBN: (纸本)9781643681351;9781643681344

cracks are an important sign of distress of concrete bridges and may reduce their service life and safety. For the case where there are stains, peeling, scratches, and uneven illumination on the surface of concrete bridges, and where it is difficult to accurately detect complete cracks, this paper proposes a new method to connect the breaks in cracks by adaptive morphological dilation based on crack direction. Most of the existing crack image detection methods attempt to achieve high detection accuracy by increasing the algorithm complexity but sacrifice real-time detection efficiency. A multiple filtering method based on a few adaptive feature thresholds is proposed to filter non-cracks and obtain a clear crack image by analyzing the morphological characteristic differences between real cracks and noise and pseudo-cracks The experimental results show that the proposed method can effectively improve the integrity of cracks, remove different noise and pseudo-cracks, without modeling, and has a higher detection accuracy and speed, which is suitable for practical engineering applications.

关键词： concrete crack detection infrared image morphology multiple feature thresholds adaptive filtering

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：