Near-surface defect detection in ultrasonic testing using domain-knowledge-informed self-supervised learning

IF 3.8 2区 物理与天体物理 Q1 ACOUSTICS
Minsu Jeon , Minseok Choi , Wonjae Choi , Jong Moon Ha , Hyunseok Oh
{"title":"Near-surface defect detection in ultrasonic testing using domain-knowledge-informed self-supervised learning","authors":"Minsu Jeon ,&nbsp;Minseok Choi ,&nbsp;Wonjae Choi ,&nbsp;Jong Moon Ha ,&nbsp;Hyunseok Oh","doi":"10.1016/j.ultras.2024.107528","DOIUrl":null,"url":null,"abstract":"<div><div>Recently, significant research efforts have been made to enhance ultrasonic testing (UT) by employing artificial intelligence (AI). However, collecting an extensive amount of labeled data across various testing environments to train the AI model poses significant challenges. Moreover, conventional UT typically focuses on detecting deep-depth defects, which limits the effectiveness of such methods in detecting near-surface defects. To this end, this paper proposes a novel near-surface defect detection method for ultrasonic testing that can be employed without collecting labeled data. We propose a self-supervised anomaly detection model that incorporates domain knowledge. First, synthetic faulty samples are generated by fusing the measured UT signals with the back-wall UT reflection signals, to simulate real faulty features. Unlike the CutPaste method used for computer vision applications, this synthesis method adds the back-wall echo signal to random locations by incorporating the physical principles of the superposition of ultrasonic signals. Next, a de-anomaly network is devised to isolate subtle defect features within the measured UT signals. The presence of defects was determined using the three-sigma rule of the mean absolute value of the residual output. The defect depth is determined by a time-of-flight calculation from the residual output. The effectiveness of the proposed method was evaluated through the UT of aluminum blocks with near-surface defects of varying depths under different surface conditions. Both qualitative and quantitative comparison studies demonstrated that the proposed method outperformed existing methods in detecting the presence and depth of near-surface defects.</div></div>","PeriodicalId":23522,"journal":{"name":"Ultrasonics","volume":"147 ","pages":"Article 107528"},"PeriodicalIF":3.8000,"publicationDate":"2024-11-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Ultrasonics","FirstCategoryId":"101","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S0041624X24002919","RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"ACOUSTICS","Score":null,"Total":0}
引用次数: 0

Abstract

Recently, significant research efforts have been made to enhance ultrasonic testing (UT) by employing artificial intelligence (AI). However, collecting an extensive amount of labeled data across various testing environments to train the AI model poses significant challenges. Moreover, conventional UT typically focuses on detecting deep-depth defects, which limits the effectiveness of such methods in detecting near-surface defects. To this end, this paper proposes a novel near-surface defect detection method for ultrasonic testing that can be employed without collecting labeled data. We propose a self-supervised anomaly detection model that incorporates domain knowledge. First, synthetic faulty samples are generated by fusing the measured UT signals with the back-wall UT reflection signals, to simulate real faulty features. Unlike the CutPaste method used for computer vision applications, this synthesis method adds the back-wall echo signal to random locations by incorporating the physical principles of the superposition of ultrasonic signals. Next, a de-anomaly network is devised to isolate subtle defect features within the measured UT signals. The presence of defects was determined using the three-sigma rule of the mean absolute value of the residual output. The defect depth is determined by a time-of-flight calculation from the residual output. The effectiveness of the proposed method was evaluated through the UT of aluminum blocks with near-surface defects of varying depths under different surface conditions. Both qualitative and quantitative comparison studies demonstrated that the proposed method outperformed existing methods in detecting the presence and depth of near-surface defects.
利用基于领域知识的自监督学习进行超声波测试中的近表面缺陷检测
最近,通过采用人工智能(AI)来增强超声波测试(UT)的研究取得了重大进展。然而,在各种测试环境中收集大量标记数据来训练人工智能模型是一项重大挑战。此外,传统的 UT 通常侧重于检测深层缺陷,这限制了此类方法在检测近表面缺陷方面的有效性。为此,本文提出了一种新颖的超声波检测近表面缺陷检测方法,该方法无需收集标记数据即可使用。我们提出了一种结合领域知识的自监督异常检测模型。首先,通过将测量到的 UT 信号与后墙 UT 反射信号融合,生成合成故障样本,以模拟真实的故障特征。与计算机视觉应用中使用的剪贴法不同,这种合成方法通过结合超声波信号叠加的物理原理,在随机位置添加后墙回波信号。接下来,设计了一个去异常网络,以隔离测量到的 UT 信号中细微的缺陷特征。利用残差输出平均绝对值的三西格玛法则确定是否存在缺陷。缺陷深度通过残差输出的飞行时间计算来确定。通过在不同表面条件下对存在不同深度近表面缺陷的铝块进行 UT,评估了所建议方法的有效性。定性和定量对比研究表明,在检测近表面缺陷的存在和深度方面,所提出的方法优于现有方法。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
求助全文
约1分钟内获得全文 求助全文
来源期刊
Ultrasonics
Ultrasonics 医学-核医学
CiteScore
7.60
自引率
19.00%
发文量
186
审稿时长
3.9 months
期刊介绍: Ultrasonics is the only internationally established journal which covers the entire field of ultrasound research and technology and all its many applications. Ultrasonics contains a variety of sections to keep readers fully informed and up-to-date on the whole spectrum of research and development throughout the world. Ultrasonics publishes papers of exceptional quality and of relevance to both academia and industry. Manuscripts in which ultrasonics is a central issue and not simply an incidental tool or minor issue, are welcomed. As well as top quality original research papers and review articles by world renowned experts, Ultrasonics also regularly features short communications, a calendar of forthcoming events and special issues dedicated to topical subjects.
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术官方微信