Automated detection of textured-surface defects using UNet-based semantic segmentation network

Enshaei, Nastaran; Ahmad, Safwan; Naderkhani, Farnoosh

doi:10.1109/icphm49022.2020.9187023

Cited by 29 publications

(13 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Many approaches use deep segmentation networks for this purpose. For instance, the authors in [12] propose an end-to-end UNet-shaped fully convolutional neural network for automated defect detection in surfaces of manufactured components. They report results over a publicly available 10 class dataset applying a realtime data augmentation approach during the training phase, achieving a mean intersection over union (IoU) of 68.35%.…”

Section: Deep Learning Based Methods For Defect Detectionmentioning

confidence: 99%

Generative Adversarial Networks to Improve the Robustness of Visual Defect Segmentation by Semantic Networks in Manufacturing Components

et al. 2021

View full text Add to dashboard Cite

This paper describes the application of Semantic Networks for the detection of defects in images of metallic manufactured components in a situation where the number of available samples of defects is small, which is rather common in real practical environments. In order to overcome this shortage of data, the common approach is to use conventional data augmentation techniques. We resort to Generative Adversarial Networks (GANs) that have shown the capability to generate highly convincing samples of a specific class as a result of a game between a discriminator and a generator module. Here, we apply the GANs to generate samples of images of metallic manufactured components with specific defects, in order to improve training of Semantic Networks (specifically DeepLabV3+ and Pyramid Attention Network (PAN) networks) carrying out the defect detection and segmentation. Our process carries out the generation of defect images using the StyleGAN2 with the DiffAugment method, followed by a conventional data augmentation over the entire enriched dataset, achieving a large balanced dataset that allows robust training of the Semantic Network. We demonstrate the approach on a private dataset generated for an industrial client, where images are captured by an ad-hoc photometric-stereo image acquisition system, and a public dataset, the Northeastern University surface defect database (NEU). The proposed approach achieves an improvement of 7% and 6% in an intersection over union (IoU) measure of detection performance on each dataset over the conventional data augmentation.

show abstract

Section: Deep Learning Based Methods For Defect Detectionmentioning

confidence: 99%

Generative Adversarial Networks to Improve the Robustness of Visual Defect Segmentation by Semantic Networks in Manufacturing Components

et al. 2021

View full text Add to dashboard Cite

show abstract

“…Similarly, the U-Net is widely used in infrastructure inspections. Enshaei et al [ 63 ] efficiently inspected textured surface defects using the U-Net. Pan et al [ 64 ] implemented an improved U-Net model to inspect sewer pipes.…”

Section: Vision-based Infrastructure Inspectionmentioning

confidence: 99%

A Review of Vision-Laser-Based Civil Infrastructure Inspection and Monitoring

Zhou

Tang

et al. 2022

Sensors

View full text Add to dashboard Cite

Structural health and construction security are important problems in civil engineering. Regular infrastructure inspection and monitoring methods are mostly performed manually. Early automatic structural health monitoring techniques were mostly based on contact sensors, which usually are difficult to maintain in complex infrastructure environments. Therefore, non-contact infrastructure inspection and monitoring techniques received increasing interest in recent years, and they are widely used in all aspects of infrastructure life, owing to their convenience and non-destructive properties. This paper provides an overview of vision-based inspection and vision–laser-based monitoring techniques and applications. The inspection part includes image-processing algorithms, object detection, and semantic segmentation. In particular, infrastructure monitoring involves not only visual technologies but also different fusion methods of vision and lasers. Furthermore, the most important challenges for future automatic non-contact inspections and monitoring are discussed and the paper correspondingly concludes with state-of-the-art algorithms and applications to resolve these challenges.

show abstract

“…In order to deal with the limited amount of ECM data and corresponding CT images, the U-net model architecture was chosen for CT image prediction since it allows a image segmentation with only few training samples [20]. It was widely applied in manufacturing for the automated analysis of visual inspections [10,7,28,17]. The U-net's bottle neck architecture forces the CNN to extract those features out of the input (ECM), thus it can best reconstruct the desired output (CT image).…”

Section: Deep Learningmentioning

confidence: 99%

How To Develop a NDT Method For Weld Inspection in Battery Cell Manufacturing Using Deep Learning

Rohkohl¹,

Kraken²,

Schönemann³

et al. 2021

Preprint

View full text Add to dashboard Cite

Battery cells are central components of electric vehicles. It is important for automotive OEM to utilize high quality battery cells to ensure high performance and safety of their vehicles. This results in the high demand for quality control measures and inspection methods in battery cell manufacturing. Particular relevant features of battery cells are welds for the internal electrical contact. Failures of these welds are often the cause for battery defects in the field and scrap during production. Consequently, there is a strong need to evaluate all welds during manufacturing. However, there is no established method which allows a quick, comprehensive, and cheap inline measurement of the weld quality. This paper presents a new eddy current based method for non-destructive testing of seam welds as well as a machine learning approach for its validation. A deep learning model has been trained on eddy current measurements to predict results from a reference inspection method, in this case computer tomography. The results prove that eddy current measurements can be used to replicate data acquired by computer tomography which means that eddy current measurements could be a suitable candidate for non-destructive 100% inline inspection. More general, this study demonstrates how machine learning may help to get deeper insights into measurement results and to validate new non-destructive testing techniques whose detailed features are yet unknown. The presented evaluation method enables understanding the capabilities and the limits of a new technique and to extract hidden features from the data. Furthermore, the usage of machine learning allows to perform these evaluations on artificial product samples with specific defects and features, which avoids the costly production physical samples.

show abstract

Automated detection of textured-surface defects using UNet-based semantic segmentation network

Cited by 29 publications

References 19 publications

Generative Adversarial Networks to Improve the Robustness of Visual Defect Segmentation by Semantic Networks in Manufacturing Components

Generative Adversarial Networks to Improve the Robustness of Visual Defect Segmentation by Semantic Networks in Manufacturing Components

A Review of Vision-Laser-Based Civil Infrastructure Inspection and Monitoring

How To Develop a NDT Method For Weld Inspection in Battery Cell Manufacturing Using Deep Learning

Contact Info

Product

Resources

About