DefectDet: A deep learning architecture for detection of defects with extreme aspect ratios in ultrasonic images

Medak, Duje; Posilović, Luka; Subašić, Marko; Budimir, Marko; Lončarić, Sven

doi:10.1016/j.neucom.2021.12.008

Cited by 25 publications

(10 citation statements)

References 43 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We developed a generative neural network for augmenting the available dataset for training the defect detection algorithms in [5]. We also developed a self-supervised anomaly detection method in [7], and a supervised learning defect detection method on B-scans in [6] and on C-scans in [9]. e data acquisition process, the ultrasound phased array transducer sends a sound acoustic radiation beams.…”

Section: Discussionmentioning

confidence: 99%

“…In [5] the same synthetic data generated by a GAN was shown to be of such high quality that human experts could not distinguish them from the real ultrasonic data. In [6] authors developed a novel convolutional neural network that outperforms all current models on ultrasonic defect detection. In [7,8] authors managed to develop a deep learning approach that utilizes only non-anomalous data for training the defect, or anomaly, detector network.…”

mentioning

confidence: 99%

See 1 more Smart Citation

Improvement Possibilities for Nuclear Power Plants Inspections by Adding Deep Learning-based Assistance Algorithms Into a Classic Ultrasound NDE Acquisition and Analysis Software

Pavlović¹,

Budimir²,

Milkovic

et al. 2023

joe

View full text Add to dashboard Cite

The safety of nuclear power plants has always been one of the most important security issues in the industry in general. Numerous standards, techniques, and tools have been developed to deal specifically with the safety of nuclear power plants – one has specialised probes, robotized systems, electronics, and software. Although seen as a mature (or slowly evolving) industry, this notion about nuclear safety is a bit misleading – the area is developing in many promising new directions. Some recent global events will speed up this development even more. On the other hand, the industry is currently going through digital transformation, which brings networking of devices, equipment, computers, and humans. This fourth industrial revolution promises speed, reliability, and efficiencies not possible up until now. In the NDE sector, new production techniques and traditional manufacturing lines are getting to be lights-out operations (near-total automation). The same is most probably going to happen with the safety inspections and quality insurance. Robotics and automation are improving worker safety and reducing human error. The well-being of inspectors working in a hazardous environment is being taken care of. Most experts agree that the digitalization of NDE offers unprecedented opportunities to the world of inspection for infrastructure safety, inspector well-being, and even product design improvements. While the community tends to agree on the value proposition of digital transformation of NDE, it also recognizes the challenges associated with such a major shift in a well-established and regulated sector. The work presented in this paper shows a part of the project that aims to develop a modular ultrasound diagnostic NDE system (consisting of exchangeable transducers, electronics, and acquisition/analysis software algorithms), for applications in hazardous environments within nuclear power plants. The paper will show how the software part of this system can reach near-total automation by implementing various deep learning algorithms as its features and, then, testing those algorithms on laboratory samples, showing encouraging results and promises of online monitoring applications. Furthermore, future general prospects of this technology are discussed, and how this technology can affect the well-being of nuclear power plant inspectors and contribute to overall plant safety.

show abstract

Section: Discussionmentioning

confidence: 99%

mentioning

confidence: 99%

Improvement Possibilities for Nuclear Power Plants Inspections by Adding Deep Learning-based Assistance Algorithms Into a Classic Ultrasound NDE Acquisition and Analysis Software

Pavlović¹,

Budimir²,

Milkovic

et al. 2023

joe

View full text Add to dashboard Cite

show abstract

“…Zhang et al used neural network technology to classify defects in ultrasound B-scan images, and proposed UFDNASNet to achieve high accuracy and processing speed detection of defects in ultrasound images 15 . D. Medak et al proposed the DefectDet model to assist detection experts in detecting and locating defects in ultrasound images, in response to the issue of errors in manually analyzing ultrasound testing data 16 . A. Karthikeyan proposed a new interpretable artificial intelligence (XAI) injection ultrasound imaging principle, which can quickly and comprehensively inspect products made of different materials, detect voids, pores, and other defects inside the workpiece 17 .…”

Section: Related Workmentioning

confidence: 99%

Promptable model for premium connection sealing surface detection and segmentation

Yang,

Fan,

Tian

et al. 2024

Fifth International Conference on Computer Vision and Computational Intelligence (CVCI 2024)

View full text Add to dashboard Cite

When applying the tubing ultrasonic testing technology to evaluate the contact condition of the metal-to-metal sealing surface of premium connection, it is necessary to judge the ultrasonic reflection amplitude image of the manual contact interface to observe whether there is sealing defect. At present, the images of phased array ultrasonic testing results usually require professionals to rely on technical knowledge to judge, the analysis has low efficiency and strong evaluation subjectivity. Therefore, there is an urgent need for an intelligent method to identify the location, range and type of sealing defects in ultrasonic images accurately and efficiently, so as assisting or replacing manual operations. Aiming at the problems of heavy reliance on data quality and inflexible segmentation effect during the process of identifying the sealing surface region using the original Mask R-CNN network, the approach described in this paper enhances the model by incorporating the Segment Anything Model(SAM) and employs prompts to guide the object detection model in generating masks that fulfill various criteria. Experiments show that the method adopted in this paper can not only correctly identify the sealing surface location, but also, compared with the original Mask R-CNN network model, it can output a segmentation mask that meets the demand according to the prompts of different segmentation criteria, and the obtained sealing surface region is closer to the theoretical segmentation region.

show abstract

“…Internal defects in materials generally cause changes in sound, light, heat, magnetism and electricity, and the NDT technology evaluates material defects by capturing real-time changes in these properties. At presently, there are various types of NDT techniques that are used for defects detection, such as pulsed eddy current testing [16][17][18], ultrasonic inspection [19][20][21][22][23], magnetic particle inspection [24,25], dye penetrant testing [26,27], and X-ray radiography [28,29]. In recent years, the rapid development of infrared thermal imagers has led to the widespread application of the infrared thermal imaging technology, and the active infrared thermography (IRT) has achieved success in many engineering practices, such as corrosion monitoring of metals and detection of internal defects in composite materials.…”

Section: Introductionmentioning

confidence: 99%

Detection of flat-bottom holes in composite materials using multi-dimensional complementary ensemble empirical mode decomposition algorithm

Zhang,

Li,

Wang

et al. 2023

J. Inst.

View full text Add to dashboard Cite

Due to high-temperature resistance, high strength, and excellent fatigue resistance, composite materials are widely used in automotive manufacturing, aerospace, infrastructure and other fields. Consequently, the demand for defect detection of composite materials is also increasing. As a non-destructive testing technique, the active infrared thermography, which can achieve full-field defect detection, is suitable for defect detection of composite materials. However, this method is susceptible to noises caused by the environment and heating sources. In order to solve the problem of the defect signal being submerged by these noises, a multi-dimensional complementary ensemble empirical mode decomposition (MCEEMD) algorithm is introduced in this paper. This method can decompose the signal into the low-frequency background noise, the high-frequency heating noise, and useful defect signals, and these noises can be easily removed to improve the contrast to noise ratio (CNR) of defect images. Based on this proposed method, a defect detection experiment on the carbon fiber reinforced plastic (CFRP) is performed in this paper, and experimental results show that the method can effectively remove environmental noise and heating noise, and it can detect 11 out of 12 defects on the CFRP sample with an average CNR of 9.107. Compared with the traditional differential absolute contrast method, this method can detect one additional small defect with the aspect ratio of 1.67 and one deep defect with a depth of 2 mm.

show abstract

DefectDet: A deep learning architecture for detection of defects with extreme aspect ratios in ultrasonic images

Cited by 25 publications

References 43 publications

Improvement Possibilities for Nuclear Power Plants Inspections by Adding Deep Learning-based Assistance Algorithms Into a Classic Ultrasound NDE Acquisition and Analysis Software

Improvement Possibilities for Nuclear Power Plants Inspections by Adding Deep Learning-based Assistance Algorithms Into a Classic Ultrasound NDE Acquisition and Analysis Software

Promptable model for premium connection sealing surface detection and segmentation

Detection of flat-bottom holes in composite materials using multi-dimensional complementary ensemble empirical mode decomposition algorithm

Contact Info

Product

Resources

About