A New GAN-Based Approach to Data Augmentation and Image Segmentation for Crack Detection in Thermal Imaging Tests

Tian, Lulu; Wang, Zidong; Liu, Weibo; Cheng, Yuhua; Alsaadi, Fuad E.; Liu, Xiaohui

doi:10.1007/s12559-021-09922-w

Cited by 25 publications

(9 citation statements)

References 57 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The proposed regularization of the generator through the addition of mean square error allowed the network to robustly distinguish the features between the signals containing noise and cracks. Similar nondestructive testing algorithms using GANs have been explored by Tian et al, 26 which explored the feasibility of data augmentation for eddy current pulsed thermography images. A penalty term was introduced to modify the traditional loss function of the GAN network to improve overall feature extraction and enhance the quality of the generated thermographic image when compared with traditional GANs.…”

Section: Introductionmentioning

confidence: 98%

Multiclass damage detection in concrete structures using a transfer learning‐based generative adversarial networks

Dunphy

Sadhu

Wang

2022

Structural Contr & Hlth

View full text Add to dashboard Cite

A large amount of the world's existing infrastructure is reaching the end of its service life, requiring intervention in the form of structural rehabilitation or replacement. A critical aspect of such asset management is the condition assessment of these structures to evaluate their existing health and dictate the scheduling and extent of required rehabilitation. It has been demonstrated that human-based manual inspections face logistical constraints and are expensive, time extensive, and subjective, depending on the knowledge of the inspection. Recently, autonomous vision-based techniques have been proposed as an alternative, more accurate method for the inspection of deteriorating structures. Convolutional neural networks (CNNs) have demonstrated state-of-the-art accuracy with respect to damage classification for concrete structures and are often implemented to process images taken from vision-based sensors such as cameras, smartphones, and drones. However, these archetypes require a large database of annotated images to train the network to an accurate level, which is not readily available for real-life structures. Moreover, CNNs are limited to the extent by which they are trained; they are often only trained for binary damage classification of a singular material model. This paper addresses these challenges of CNNs through the application of a generative adversarial network (GANs) for multiclass damage detection of concrete structures. The proposed GAN is trained using the SDNET2018 dataset to detect cracking, spalling, pitting, and construction joints in concrete surfaces. Moreover, transfer learning is implemented to transfer the learned features of the GAN to a CNN architecture to allow for accurate image classification. It is concluded that, for a 0%-30% reduction in the amount of labeled data used, the proposed GAN method has comparable accuracy to traditional CNNs.

show abstract

Section: Introductionmentioning

confidence: 98%

Multiclass damage detection in concrete structures using a transfer learning‐based generative adversarial networks

Dunphy

Sadhu

Wang

2022

Structural Contr & Hlth

View full text Add to dashboard Cite

show abstract

“…Recently, with the fast advancement of science and technology, image enhancement technology has been applied in the field of medical diagnosis. In the process of imaging, medical images will be interfered by external factors such as imaging equipment and temperature, resulting in low contrast and noise pollution of medical images, which will directly affect the accuracy of post‐processing operations such as image segmentation 1,2 . For this reason, image enhancement is extremely essential for practical applications and is a pivotal step in image pre‐processing.…”

Section: Introductionmentioning

confidence: 99%

“…In the process of imaging, medical images will be interfered by external factors such as imaging equipment and temperature, resulting in low contrast and noise pollution of medical images, which will directly affect the accuracy of post-processing operations such as image segmentation. 1,2 For this reason, image enhancement is extremely essential for practical applications and is a pivotal step in image pre-processing. At present, there are mainly two kinds of image enhancement methods, namely spatial domain-based enhancement and transform domain-based enhancement.…”

Section: Introductionmentioning

confidence: 99%

Medical image contrast enhancement based on improved sparrow search algorithm

Fan

Sun

Yin

et al. 2022

Int J Imaging Syst Tech

View full text Add to dashboard Cite

The collection or transmission of medical images is often disturbed by various factors, such as insufficient brightness and noise pollution, which will result in the deterioration of image quality and significantly affect the clinical diagnosis.To improve the quality of medical images, a contrast enhancement method based on improved sparrow search algorithm is proposed in this paper. The method is divided into two steps to enhance the medical images. First, a new transform function is introduced to improve the brightness or contrast of medical images, and two parameters in the transform function are optimized by the improved sparrow search algorithm. Second, adaptive histogram equalization method with contrast limited is used to equalize the result image of the previous step to make the pixel distribution of the image more uniform. Finally, a large number of experiments and qualitative and quantitative analyses were conducted on the common data sets. The analysis results demonstrate that the presented approach outperforms some existing medical image processing approaches.

show abstract

“…It has better results for feature extraction and recognition. A large amount of image data can often be generated by eddy current and ultrasonic inspection techniques, which fits well with neural networks (Tian et al, 2021). The features of rail surface images are extracted by neural networks (Han et al, 2021) or by combining neural networks with saliency cueing methods (Lu et al, 2020), both of which perform well for automated identification of rail damage.…”

Section: Introductionmentioning

confidence: 99%

An improved convolutional neural network for convenient rail damage detection

2022

View full text Add to dashboard Cite

The long-term operation of a railroad usually leads to defects in its rails, axles, fasteners, etc. These problems directly affect the safety of the rail system. Therefore, it is important to ensure the health of key railroad structures. In this paper, a deep learning-based rail damage identification method is established by analyzing the rail vibration signals collected with piezoelectric ceramic pads. The multiple features of vibration signals are combined and then reconstructed into grayscale maps as the inputs of the model. The key information of the grayscale maps is extracted using neural networks. The idea of pre-convolution is used to solve the problem that the model pays more attention to certain features due to the different input sizes and the implied weights of the features. Finally, the performance of the three convolutional neural networks (CNN) in rail damage detection is evaluated and compared. The results show that the CNN with pre-convolution and Residual structure has better recognition for the presence of rail damage than other methods.

show abstract

A New GAN-Based Approach to Data Augmentation and Image Segmentation for Crack Detection in Thermal Imaging Tests

Cited by 25 publications

References 57 publications

Multiclass damage detection in concrete structures using a transfer learning‐based generative adversarial networks

Multiclass damage detection in concrete structures using a transfer learning‐based generative adversarial networks

Medical image contrast enhancement based on improved sparrow search algorithm

An improved convolutional neural network for convenient rail damage detection

Contact Info

Product

Resources

About