A Lightweight Network with Dual Encoder and Cross Feature Fusion for Cement Pavement Crack Detection

Automatic crack detection of cement pavement chiefly benefits from the rapid development of deep learning,with convolutional neural networks(CNN)playing an important role in this field.However,as the performance of crack detection in cement pavement improves,the depth and width of the network structure are significantly increased,which necessitates more computing power and storage space.This limitation hampers the practical implementation of crack detection models on various platforms,particularly portable devices like small mobile devices.To solve these problems,we propose a dual-encoder-based network architecture that focuses on extracting more comprehensive fracture feature information and combines cross-fusion modules and coordinated attention mechanisms formore efficient feature fusion.Firstly,we use small channel convolution to construct shallow feature extractionmodule(SFEM)to extract low-level feature information of cracks in cement pavement images,in order to obtainmore information about cracks in the shallowfeatures of images.In addition,we construct large kernel atrous convolution(LKAC)to enhance crack information,which incorporates coordination attention mechanism for non-crack information filtering,and large kernel atrous convolution with different cores,using different receptive fields to extract more detailed edge and context information.Finally,the three-stage feature map outputs from the shallow feature extraction module is cross-fused with the two-stage feature map outputs from the large kernel atrous convolution module,and the shallow feature and detailed edge feature are fully fused to obtain the final crack prediction map.We evaluate our method on three public crack datasets:DeepCrack,CFD,and Crack500.Experimental results on theDeepCrack dataset demonstrate the effectiveness of our proposed method compared to state-of-the-art crack detection methods,which achieves Precision(P)87.2%,Recall(R)87.7%,and F-score(F1)87.4%.Thanks to our lightweight crack detectionmodel,the parameter count of the model in real-world detection scenarios has been significantly reduced to less than 2M.This advancement also facilitates technical support for portable scene detection.

A Rapid Crack Detection Technique Based on Attention for Intelligent M&O of Cross-Sea Bridge

Rapid and accurate segmentation of structural cracks is essential for ensuring the quality and safety of engineering projects.In practice,however,this task faces the challenge of finding a balance between detection accuracy and efficiency.To alleviate this problem,a lightweight and efficient real-time crack segmentation framework was developed.Specifically,in the network model system based on an encoding-decoding structure,the encoding network is equipped with packet convolution and attention mechanisms to capture features of different visual scales in layers,and in the decoding process,we also introduce a fusion module based on spatial attention to effectively aggregate these hierarchical features.Codecs are connected by pyramid pooling model(PPM)filtering.The results show that the crack segmentation accuracy and real-time operation capability larger than 76%and 15 fps,respectively,are validated by three publicly available datasets.These wide-ranging results highlight the potential of the model for the intelligent O&M for cross-sea bridge.

Automatic Pavement Crack Detection Based on Octave Convolution Neural Network with Hierarchical Feature Learning

Automatic pavement crack detection plays an important role in ensuring road safety.In images of cracks,information about the cracks can be conveyed through high-frequency and low-fre-quency signals that focus on fine details and global structures,respectively.The output features obtained from different convolutional layers can be combined to represent information about both high-frequency and low-frequency signals.In this paper,we propose an encoder-decoder framework called octave hierarchical network(Octave-H),which is based on the U-Network(U-Net)architec-ture and utilizes an octave convolutional neural network and a hierarchical feature learning module for performing crack detection.The proposed octave convolution is capable of extracting multi-fre-quency feature maps,capturing both fine details and global cracks.We propose a hierarchical feature learning module that merges multi-frequency-scale feature maps with different levels(high and low)of octave convolutional layers.To verify the superiority of the proposed Octave-H,we employed the CrackForest dataset(CFD)and AigleRN databases to evaluate this method.The experimental results demonstrate that Octave-H outperforms other algorithms with satisfactory performance.

Crack detection of reinforced concrete bridge using video image

With the digital image technology,a crack detection method of reinforced concrete bridge was studied for the performance assessment.The effects including the image gray level,pixel rate,noise filter,and edge detection were analyzed considering cracks qualities.A computer program was developed by visual C++6.0 programming language to detect the cracks,which was tested by 15cases of bridge video images.The results indicate that the relative error is within 6%for cracks larger than 0.3 mm cracks and it is less than 10%for crack width between 0.2 mm and 0.3 mm.In addition,for the crack below 0.1 mm,the relative error is more than30%because the bridge is in safe stage and it is very difficult to detect the actual width of crack.

Piezoelectric-based Crack Detection Techniques of Concrete Structures:Experimental Study

Feasibility of a wave propagation-based active crack detection technique for nondestructive evaluations （NDE） of concrete structures with surface bonded and embedded piezoelectric-ceramic （PZT） patches was studied. At first, the wave propagation mechanisms in concrete were analyzed. Then, an active sensing system with integrated actuators/sensors was constructed. One PZT patch was used as an actuator to generate high frequency waves, and the other PZT patches were used as sensors to detect the propagating wave. Scattered wave signals from the damage can be obtained by subtracting the baseline signal of the intact structure from the recorded signal of the damaged structure. In the experimental study, progressive cracked damage inflicted artificially on the plain concrete beam is assessed by using both lateral and thickness modes of the PZT patches. The results indicate that with the increasing number and severity of cracks, the magnitude of the sensor output decreases for the surface bonded PZT patches, and increases for the embedded PZT patches.

Pavement Cracks Coupled With Shadows:A New Shadow-Crack Dataset and A Shadow-Removal-Oriented Crack Detection Approach

Automatic pavement crack detection is a critical task for maintaining the pavement stability and driving safety.The task is challenging because the shadows on the pavement may have similar intensity with the crack,which interfere with the crack detection performance.Till to the present,there still lacks efficient algorithm models and training datasets to deal with the interference brought by the shadows.To fill in the gap,we made several contributions as follows.First,we proposed a new pavement shadow and crack dataset,which contains a variety of shadow and pavement pixel size combinations.It also covers all common cracks(linear cracks and network cracks),placing higher demands on crack detection methods.Second,we designed a two-step shadow-removal-oriented crack detection approach:SROCD,which improves the performance of the algorithm by first removing the shadow and then detecting it.In addition to shadows,the method can cope with other noise disturbances.Third,we explored the mechanism of how shadows affect crack detection.Based on this mechanism,we propose a data augmentation method based on the difference in brightness values,which can adapt to brightness changes caused by seasonal and weather changes.Finally,we introduced a residual feature augmentation algorithm to detect small cracks that can predict sudden disasters,and the algorithm improves the performance of the model overall.We compare our method with the state-of-the-art methods on existing pavement crack datasets and the shadow-crack dataset,and the experimental results demonstrate the superiority of our method.

Crack detection using a frequency response function in offshore platforms

Structural cracks can change the frequency response function (FRF) of an offshore platform. Thus, FRF shifts can be used to detect cracks. When a crack at a specific location and magnitude occurs in an offshore structure, changes in the FRF can be measured. In this way, shifts in FRF can be used to detect cracks. An experimental model was constructed to verify the FRF method. The relationship between FRF and cracks was found to be non-linear. The effect of multiple cracks on FRF was analyzed, and the shift due to multiple cracks was found to be much more than the summation of FRF shifts due to each of the cracks. Then the effects of noise and changes in the mass of the jacket on FRF were evaluated. The results show that significant damage to a beam can be detected by dramatic changes in the FRF, even when 10% random noise exists. FRF can also be used to approximately locate the breakage, but it can neither be efficiently used to predict the location of breakage nor the existence of small hairline cracks. The FRF shift caused by a 7% mass change is much less than the FRF shift caused by the breakage of any beam, but is larger than that caused by any early cracks.

A Study on Crack Detection with Modal Parameters of A Jacket Platform

Crack detection procedures by different modal parameters are analyzed for identifying a crack and its location and magnitude in a jacket platform. The first ten natural frequencies and modal shapes of the jacket models are obtained by numerical experiments based on NASTRAN Code. A crack at different locations and of different magnitudes is imposed in the model at the underwater beams. Then, the modal evaluation parameters are calculated numerically, to illustrate the evaluation of modal parameter criteria used in jacket crack detection. The sensitivities of different modal parameters to different cracks are analyzed. A new technique is presented for predicting the approximate location of a breakage in the absence of the data of an intact model. This method can be used to detect a crack in underwater members by use of incomplete mode shapes of the top members of the jacket.

AN IMPEDANCE ANALYSIS FOR CRACK DETECTION IN THE TIMOSHENKO BEAM BASED ON THE ANTI-RESONANCE TECHNIQUE

An alternative technique for crack detection in a Timoshenko beam based on the first anti-resonant frequency is presented in this paper. Unlike the natural frequency, the anti-resonant frequency is a local parameter rather than a global parameter of structures, thus the proposed technique can be used to locate the structural defects. An impedance analysis of a cracked beam stimulated by a harmonic force based on the Timoshenko beam formulation is investigated. In order to characterize the local discontinuity due to cracks, a rotational spring model based on fracture mechanics is proposed to model the crack. Subsequently, the proposed method is verified by a numerical example of a simply-supported beam with a crack. The effect of the crack size on the anti-resonant frequency is investigated. The position of the crack of the simply-supported beam is also determined by the anti-resonance technique. The proposed technique is further applied to the ＂contaminated＂ anti-resonant frequency to detect crack damage, which is obtained by adding 1-3% noise to the calculated data. It is found that the proposed technique is effective and free from the environment noise. Finally, an experimental study is performed, which further verifies the validity of the proposed crack identification technique.

Vision-based fatigue crack detection using global motion compensation and video feature tracking

Fatigue cracks that develop in civil infrastructure such as steel bridges due to repetitive loads pose a major threat to structural integrity.Despite being the most common practice for fatigue crack detection,human visual inspection is known to be labor intensive,time-consuming,and prone to error.In this study,a computer vision-based fatigue crack detection approach using a short video recorded under live loads by a moving consumer-grade camera is presented.The method detects fatigue crack by tracking surface motion and identifies the differential motion pattern caused by opening and closing of the fatigue crack.However,the global motion introduced by a moving camera in the recorded video is typically far greater than the actual motion associated with fatigue crack opening/closing,leading to false detection results.To overcome the challenge,global motion compensation(GMC)techniques are introduced to compensate for camera-induced movement.In particular,hierarchical model-based motion estimation is adopted for 2D videos with simple geometry and a new method is developed by extending the bundled camera paths approach for 3D videos with complex geometry.The proposed methodology is validated using two laboratory test setups for both in-plane and out-of-plane fatigue cracks.The results confirm the importance of motion compensation for both 2D and 3D videos and demonstrate the effectiveness of the proposed GMC methods as well as the subsequent crack detection algorithm.

Automated Pavement Crack Detection Using Deep Feature Selection and Whale Optimization Algorithm

Pavement crack detection plays a crucial role in ensuring road safety and reducing maintenance expenses.Recent advancements in deep learning(DL)techniques have shown promising results in detecting pavement cracks;however,the selection of relevant features for classification remains challenging.In this study,we propose a new approach for pavement crack detection that integrates deep learning for feature extraction,the whale optimization algorithm(WOA)for feature selection,and random forest(RF)for classification.The performance of the models was evaluated using accuracy,recall,precision,F1 score,and area under the receiver operating characteristic curve(AUC).Our findings reveal that Model 2,which incorporates RF into the ResNet-18 architecture,outperforms baseline Model 1 across all evaluation metrics.Nevertheless,our proposed model,which combines ResNet-18 with both WOA and RF,achieves significantly higher accuracy,recall,precision,and F1 score compared to the other two models.These results underscore the effectiveness of integrating RF and WOA into ResNet-18 for pavement crack detection applications.We applied the proposed approach to a dataset of pavement images,achieving an accuracy of 97.16%and an AUC of 0.984.Our results demonstrate that the proposed approach surpasses existing methods for pavement crack detection,offering a promising solution for the automatic identification of pavement cracks.By leveraging this approach,potential safety hazards can be identified more effectively,enabling timely repairs and maintenance measures.Lastly,the findings of this study also emphasize the potential of integrating RF and WOA with deep learning for pavement crack detection,providing road authorities with the necessary tools to make informed decisions regarding road infrastructure maintenance.

Gated Fusion Based Transformer Model for Crack Detection on Wind Turbine Blade

Harsh working environments and wear between blades and other unit components can easily lead to cracks and damage on wind turbine blades.The cracks on the blades can endanger the shafting of the generator set,the tower and other components,and even cause the tower to collapse.To achieve high-precision wind blade crack detection,this paper proposes a crack fault-detection strategy that integratesGated ResidualNetwork(GRN),a fusionmodule and Transformer.Firstly,GRNcan reduce unnecessary noisy inputs that could negatively impact performancewhile preserving the integrity of feature information.In addition,to gain in-depth information about the characteristics of wind turbine blades,a fusionmodule is suggested to implement the information fusion of wind turbine features.Specifically,each fan feature ismapped to a one-dimensional vector with the same length,and all one-dimensional vectors are concatenated to obtain a two-dimensional vector.And then,in the fusion module,the information fusion of the same characteristic variables in the different channels is realized through the Channel-mixing MLP,and the information fusion of different characteristic variables in the same channel is realized through the ColumnmixingMLP.Finally,the fused feature vector is input into the Transformer for feature learning,which enhances the influence of important feature information and improves the model’s anti-noise ability and classification accuracy.Extensive experimentswere conducted on the wind turbine supervisory control and data acquisition(SCADA)data froma domesticwind field.The results show that compared with other state-of-the-artmodels,including XGBoost,LightGBM,TabNet,etc.,the F1-score of proposed gated fusion based Transformer model can reach 0.9907,which is 0.4%-2.09% higher than the comparedmodels.Thismethod provides amore reliable approach for the condition detection and maintenance of fan blades in wind farms.

Steam turbine rotor crack detection using sifting process of EMD and B-spline wavelet on the interval element model

A high-precision identification method for steam turbine rotor crack is presented. By providing me nrst three measured natural frequencies, contours for the specified natural frequency are plotted in the same coordi- nate, and the intersection of the three curves predicts the crack location and size. The cracked rotor system is mod- eled using B-spline wavelet on the interval （BSWI） finite element method, and a method based on empirical mode decomposition （EMD） and Laplace wavelet is implemented to improve the identification precision of the first three measured natural frequencies. Compared with the classical nondestructive testing, the presented method shows its effectiveness and reliability. It is feasible to apply this method to the online health monitoring for rotor structure.

Research on Infrared Image Fusion Technology Based on Road Crack Detection

This study aimed to propose road crack detection method based on infrared image fusion technology.By analyzing the characteristics of road crack images,this method uses a variety of infrared image fusion methods to process different types of images.The use of this method allows the detection of road cracks,which not only reduces the professional requirements for inspectors,but also improves the accuracy of road crack detection.Based on infrared image processing technology,on the basis of in-depth analysis of infrared image features,a road crack detection method is proposed,which can accurately identify the road crack location,direction,length,and other characteristic information.Experiments showed that this method has a good effect,and can meet the requirement of road crack detection.

Statistical Features and Traditional SA-SVM Classification Algorithm for Crack Detection

In recent years, the interest in damage identification of structural components through innovative techniques has grown significantly. Damage identification has always been a crucial concern in quality assessment and load capacity rating of infrastructure. In this regard, researchers focus on proposing efficient tools to identify the damages in early stages to prevent the sudden failure in structural components, ensuring the public safety and reducing the asset management costs. The sensing technologies along with the data analysis through various techniques and machine learning approaches have been the area of interest for these innovative techniques. The purpose of this research is to develop a robust method for automatic condition assessment of real-life concrete structures for the detection of relatively small cracks at early stages. A damage identification algorithm is proposed using the hybrid approaches to analyze the sensors data. The data obtained from transducers mounted on concrete beams under static loading in laboratory. These data are used as the input parameters. The method relies only on the measured time responses. After filtering and normalization of the data, the damage sensitive statistical features are extracted from the signals and used as the inputs of Self-Advising Support Vector Machine (SA-SVM) for the classification purpose in civil Engineering area. Finally, the results are compared with traditional methods to investigate the feasibility of the hybrid proposed algorithm. It is demonstrated that the presented method can reliably detect the crack in the structure and thereby enable the real-time infrastructure health monitoring.

Detection and Genesis Analysis of Cracks in Prestressed Box Girder of a Certain Project

This paper introduces a crack detection example of the prestressed box girder in a certain project. The morphology of the box girder cracks was surveyed and mapped. The length, width, and depth of the cracks were inspected, and the strength and reinforcement configuration of the components were tested. The test results indicate that the strength and reinforcement configuration of the inspected components meet the design requirements. The crack at the end of the top plate of the box girder is a local compressive crack at the anchorage end. The width and length of the crack on the bottom surface of the top plate are not significant, and the depth is relatively shallow. Judging from the crack morphology, this crack is identified as a temperature crack. Additionally, based on the treatment measures for cracks of different widths, the treatment measures for the cracks of the components in this project are derived, providing a reference basis for similar projects in the future.

Real-Time Detection of Cracks on Concrete Bridge Decks Using Deep Learning in the Frequency Domain

This paper presents a vision-based crack detection approach for concrete bridge decks using an integrated one-dimensional convolutional neural network(1D-CNN)and long short-term memory(LSTM)method in the image frequency domain.The so-called 1D-CNN-LSTM algorithm is trained using thousands of images of cracked and non-cracked concrete bridge decks.In order to improve the training efficiency,images are first transformed into the frequency domain during a preprocessing phase.The algorithm is then calibrated using the flattened frequency data.LSTM is used to improve the performance of the developed network for long sequence data.The accuracy of the developed model is 99.05%,98.9%,and 99.25%,respectively,for training,validation,and testing data.An implementation framework is further developed for future application of the trained model for large-scale images.The proposed 1D-CNN-LSTM method exhibits superior performance in comparison with existing deep learning methods in terms of accuracy and computation time.The fast implementation of the 1D-CNN-LSTM algorithm makes it a promising tool for real-time crack detection.

Application of signal processing and support vector machine to transverse cracking detection in asphalt pavement

Vibration-based pavement condition(roughness and obvious anomalies)monitoring has been expanding in road engineering.However,the indistinctive transverse cracking has hardly been considered.Therefore,a vehicle-based novel method is proposed for detecting the transverse cracking through signal processing techniques and support vector machine(SVM).The vibration signals of the car traveling on the transverse-cracked and the crack-free sections were subjected to signal processing in time domain,frequency domain and wavelet domain,aiming to find indices that can discriminate vibration signal between the cracked and uncracked section.These indices were used to form 8 SVM models.The model with the highest accuracy and F1-measure was preferred,consisting of features including vehicle speed,range,relative standard deviation,maximum Fourier coefficient,and wavelet coefficient.Therefore,a crack and crack-free classifier was developed.Then its feasibility was investigated by 2292 pavement sections.The detection accuracy and F1-measure are 97.25%and 85.25%,respectively.The cracking detection approach proposed in this paper and the smartphone-based detection method for IRI and other distress may form a comprehensive pavement condition survey system.

A review of deep learning methods for pixel-level crack detection

Cracks are a major sign of aging transportation infrastructure.The detection and repair of cracks is the key to ensuring the overall safety of the transportation infrastructure.In recent years,due to the remarkable success of deep learning(DL)in the field of crack detection,many researches have been devoted to developing pixel-level crack image segmentation(CIS)models based on DL to improve crack detection accuracy,but as far as we know there is no review of DL-based CIS methods yet.To address this gap,we present a comprehensive thematic survey of DL-based CIS techniques.Our review offers several contributions to the CIS area.First,more than 40 papers of journal or top conference most published in the last three years are identified and collected based on the systematic literature review method.Second,according to the backbone network architecture of the models proposed in them,they are grouped into 10 topics:FCN,U-Net,encoder-decoder model,multi-scale,attention mechanism,transformer,two-stage detection,multi-modal fusion,unsupervised learning and weakly supervised learning,to be reviewed.Meanwhile,our survey focuses on discussing strengths and limitations of the models in each topic so as to reveal the latest research progress in the CIS field.Third,publicly accessible data sets,evaluation metrics,and loss functions that can be used for pixel-level crack detection are systematically introduced and summarized to facilitate researchers to select suitable components according to their own research tasks.Finally,we discuss six common problems and existing solutions to them in the field of DL-based CIS,and then suggest eight possible future research directions in this field.

Feature analysis of acoustic emission sources for rail crack detection by the finite element method

The safety of rail is very important for the development of high speed railway, and it is necessary to investigate the features of inner cracks in rail. In order to obtain the features of Acoustic Emission （AE） sources of inner cracks in rail, AE sources with different types, depths and propagation distances are examined for crack in rail. The finite element method is utilized to model the rail with cracks and the results of experiment demonstrate the effectiveness of this model. Wavelet transform and Rayleigh-Lamb equations are utilized to extract the features of crack AE sources. The results illustrate that the intensity ratio among AE modes can identify the AE source types and the AE sources with different frequencies in rail. There are uniform AE mode features existing in the AE signals from AE sources in rail web, however AE signals from AE sources in rail head and rail base have the complex and unstable AE modes. Different AE source types have the different propagation features in rail. It is helpful to understand the rail cracks and detect the rail cracks based on the AE technique.