An integrated approach to automatic pixel-level crack detection and quantification of asphalt pavement

Ji, Ankang; Xue, Xiaolong; Wang, Yuna; Luo, Xiaowei; Xue, Weirui

doi:10.1016/j.autcon.2020.103176

Cited by 183 publications

(70 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Many studies have been proposed for the expansion of medical image data to increase the accuracy of deep learning (Bowles et al, 2018;Chang et al, 2018;Frid-Adar, Diamant et al, 2018;Han et al, 2018;M. Hu & Li, 2019;Kiyasseh et al, 2020;Leming et al, 2020;Qi et al, 2020). For example, Rashid et al (2019) proposed a GAN for obtaining realistic looking dermoscopic images on the skin lesion classification task.…”

Section: F I G U R E 1 Different Image Sources For Pavement Distress Detectionmentioning

confidence: 99%

Cross‐scene pavement distress detection by a novel transfer learning framework

Che

Liu

et al. 2021

Computer aided Civil Eng

View full text Add to dashboard Cite

Deep learning has achieved promising results in pavement distress detection. However, the training model's effectiveness varies according to the data and scenarios acquired by different camera types and their installation positions. It is time consuming and labor intensive to recollect labeled data and retrain a new model every time the scene changes. In this paper, we propose a transfer learning pipeline to address this problem, which enables a distress detection model to be applied to other untrained scenarios. The framework consists of two main components: data transfer and model transfer. The former trains a generative adversarial network to transfer existing image data into a new scene style. Then, attentive CutMix and image melding are applied to insert distress annotations to synthesize the new scene's labeled data. After data expansion, the latter step transfers the feature extracted by the existing model to the detection application of the new scene through domain adaptation. The effects of varying degrees of knowledge transfer are also discussed. The proposed method is evaluated on two data sets from two different scenes with more than 40,000 images totally. This method can reduce the demand for training data by at least 25% when the model is applied in a new scene. With the same number of training images, the proposed method can improve the model accuracy by 26.55%.

show abstract

Section: F I G U R E 1 Different Image Sources For Pavement Distress Detectionmentioning

confidence: 99%

Cross‐scene pavement distress detection by a novel transfer learning framework

Che

Liu

et al. 2021

Computer aided Civil Eng

View full text Add to dashboard Cite

show abstract

“…The morphological thinning and pruning process was used to find the average thickness and length of the cracks. Fast Parallel Thinning (FPT) technique and Adaptive skeletonisation algorithms were used to quantify the detected cracks in pixel level (Ji et al, 2020; Yang et al, 2018). Initially, the skeleton of the cracks were extracted from the output image by discarding all contour points except those belonging to the skeleton.…”

Section: Crack Quantificationmentioning

confidence: 99%

Deep learning based thermal crack detection on structural concrete exposed to elevated temperature

Andrushia

Anand

Lublóy

et al. 2021

Advances in Structural Engineering

View full text Add to dashboard Cite

Health monitoring of concrete including, detecting defects such as cracking, spalling on fire affected concrete structures plays a vital role in the maintenance of reinforced cement concrete structures. However, this process mostly uses human inspection and relies on subjective knowledge of the inspectors. To overcome this limitation, a deep learning based automatic crack detection method is proposed. Deep learning is a vibrant strategy under computer vision field. The proposed method consists of U-Net architecture with an encoder and decoder framework. It performs pixel wise classification to detect the thermal cracks accurately. Binary Cross Entropy (BCA) based loss function is selected as the evaluation function. Trained U-Net is capable of detecting major thermal cracks and minor thermal cracks under various heating durations. The proposed, U-Net crack detection is a novel method which can be used to detect the thermal cracks developed on fire exposed concrete structures. The proposed method is compared with the other state-of-the-art methods and found to be accurate with 78.12% Intersection over Union (IoU).

show abstract

“…To prevent overfitting, the study was GPS environment Good conducted using two strategies: (a) the regularization term was introduced in the loss function to restrict some parameters in the loss and thus improve the generalization capability of the model; (b) the regularization hyperparameter of weight decay was introduced in the loss function, and the weight decay value assigned was 0.00004. 45 The training process was carried out on a workstation with an NVIDIA Quadro P4000 8 GB GPU, running in a virtual environment established by Anaconda. The training and validation process took approximately 1.5 h, and DeepLabv3+ achieved a mean intersection over union (MIoU) of 0.7875.…”

Section: Crack Detection Using Deeplabv3+mentioning

confidence: 99%

“…Due to the limited number of samples, the crack type was not distinguished in this experiment; the training process can be regarded as a classification scheme with two possible outputs: “target” and “background.” Therefore, the number of classification schemes and the weight of “target” and “background” were modified for preventing class imbalance in the source code. To prevent overfitting, the study was conducted using two strategies: (a) the regularization term was introduced in the loss function to restrict some parameters in the loss and thus improve the generalization capability of the model; (b) the regularization hyperparameter of weight decay was introduced in the loss function, and the weight decay value assigned was 0.00004 45 …”

Section: Field Experimentsmentioning

confidence: 99%

Image‐based road crack risk‐informed assessment using a convolutional neural network and an unmanned aerial vehicle

Xue

Wang

et al. 2021

Struct Control Health Monit

Self Cite

View full text Add to dashboard Cite

Rapid crack assessment is widely thought to be critical for monitoring and maintaining roads in appropriate conditions. In this paper, a novel crackaffected risk-informed assessment framework is proposed for the monitoring and maintenance of roads. The framework includes five steps: data collection, crack detection, crack location extraction, crack real-size calculation, and risklevel assessment. To support the framework, an unmanned aerial vehicle (UAV) is used to monitor roads and collect data. A state-of-the-art semantic segmentation network, DeepLabv3+, is also applied to detect cracks. Based on the height and pixel statistics of the detected crack, the real size of the crack can be calculated using the pixel-physical conversion coefficient equation. This is followed by a risk-informed assessment of the road identifying the location of the crack for maintenance priority determination. Data collection and experiments using a UAV were performed on a real road to verify the feasibility and effectiveness of the proposed method. It was found that the proposed method was not only able to determine the real size of the cracks from the collected images but also able to determine their risk levels. In summary, this method presents a convenient and reliable solution for risk-informed road crack assessment that can be employed in practical applications.

show abstract

An integrated approach to automatic pixel-level crack detection and quantification of asphalt pavement

Cited by 183 publications

References 29 publications

Cross‐scene pavement distress detection by a novel transfer learning framework

Cross‐scene pavement distress detection by a novel transfer learning framework

Deep learning based thermal crack detection on structural concrete exposed to elevated temperature

Image‐based road crack risk‐informed assessment using a convolutional neural network and an unmanned aerial vehicle

Contact Info

Product

Resources

About