ABSTRACTThe operation and maintenance of the background system is always an important link to ensure the system’s high availability. With the increasing number of background systems, their operation, and maintenance have to develop from the initial huge-crowd strategy to the direction of intelligence. The key to intelligent operation and maintenance is the abnormal detection of key performance indicators (KPI), such as CPU utilisation. However, the existing KPI anomaly detection algorithms not only cannot select the dynamic threshold under the non-parametric methods but also have no false-positive correction mechanism to correct the false alarms. In order to overcome the above shortcomings, this work proposes a dynamic self-correcting Key Performance Indicator (KPI) anomaly detection algorithm, hereafter referred to as DSCAD. To the best of our knowledge, in the field of KPI anomaly detection, the DSCAD algorithm is the first dynamic threshold algorithm that does not rely on the assumption of normal distribution. Compared with the existing KPI anomaly detection methods, the F-score of the DSCAD algorithm increased by 3% and had the best performance.KEYWORDS: KPI anomaly detectiondynamic threshold selectionfalse-positive correction Disclosure statementNo potential conflict of interest was reported by the authors.Additional informationFundingThis work was supported in part by the National Natural Science Foundation of China under Grant 62072319; the Sichuan Science and Technology Program under Grant 2023YFQ0022 and 2022YFG0041; the Luzhou Science and Technology Innovation R&D Program (No. 2022CDLZ-6)
{"title":"Dynamic self-correcting key performance indicator anomaly detection algorithm","authors":"Yufang Sun, Shanghua Gao, Hongxiu Lin, Fenglin Liu, Bin Xing, Bing Guo","doi":"10.1080/10589759.2023.2273998","DOIUrl":"https://doi.org/10.1080/10589759.2023.2273998","url":null,"abstract":"ABSTRACTThe operation and maintenance of the background system is always an important link to ensure the system’s high availability. With the increasing number of background systems, their operation, and maintenance have to develop from the initial huge-crowd strategy to the direction of intelligence. The key to intelligent operation and maintenance is the abnormal detection of key performance indicators (KPI), such as CPU utilisation. However, the existing KPI anomaly detection algorithms not only cannot select the dynamic threshold under the non-parametric methods but also have no false-positive correction mechanism to correct the false alarms. In order to overcome the above shortcomings, this work proposes a dynamic self-correcting Key Performance Indicator (KPI) anomaly detection algorithm, hereafter referred to as DSCAD. To the best of our knowledge, in the field of KPI anomaly detection, the DSCAD algorithm is the first dynamic threshold algorithm that does not rely on the assumption of normal distribution. Compared with the existing KPI anomaly detection methods, the F-score of the DSCAD algorithm increased by 3% and had the best performance.KEYWORDS: KPI anomaly detectiondynamic threshold selectionfalse-positive correction Disclosure statementNo potential conflict of interest was reported by the authors.Additional informationFundingThis work was supported in part by the National Natural Science Foundation of China under Grant 62072319; the Sichuan Science and Technology Program under Grant 2023YFQ0022 and 2022YFG0041; the Luzhou Science and Technology Innovation R&D Program (No. 2022CDLZ-6)","PeriodicalId":49746,"journal":{"name":"Nondestructive Testing and Evaluation","volume":" 27","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-11-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135240837","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-11-08DOI: 10.1080/10589759.2023.2273999
Juanjuan Li, Anhong Wang
ABSTRACTTo take advantage of ultrasonic-based non-destructive testing (NDT) and data-driven intelligent defect diagnosis, the current study proposes a feature tensor classifier based on multi-source ultrasonic fusion, to enhance the defect diagnosis adaptability and reliability for gas-insulated switchgear (GIS) basin insulator. First, multi-source ultrasonic signals are acquired by finite element modelling (FEM), describing the healthy states of the GIS basin insulator completely. Second, time of flight (Tof)-featured tensors are expressed by wavelet transform (WT), and used to create the datasets. Third, a deep learning-based feature tensor classifier is proposed, and concerned training, validation, and testing processes are carried out. Finally, the effectiveness of feature tensor extraction is evaluated, and the anti-noise performance of the Tof-featured tensor classifier is verified. The main contributions indicate that the Tof-featured tensor classifier can realise excellent diagnosis performance, the average accuracy is, respectively, 90.53%, 99.75%, and 100% in training, validation, and testing sets, while the signal tensor classifier has poor performance. In addition, three other noised datasets are applied, and the result shows that the anti-noise performance of the Tof-featured tensor classifier is feasible, when SNR is greater than 1 dB.KEYWORDS: GIS basin insulatorintelligent fault diagnosismulti-source ultrasonic fusiontof-featured tensorconvolution neural network AcknowledgmentsMy heartfelt thanks are due to Prof. Han for his academic supervision and personal support. This work was supported by the Scientific and Technological Innovation Programs of Higher Education Institutions in Shanxi (Grant No. STIP2020L0699), Fund program of Key Laboratory of Signal Capturing & Processing, North University in Shanxi (Grant No. ISPT2020-8).Disclosure statementNo potential conflict of interest was reported by the author(s).
{"title":"Intelligent defect diagnosis of GIS basin insulator via multi-source ultrasonic fusion","authors":"Juanjuan Li, Anhong Wang","doi":"10.1080/10589759.2023.2273999","DOIUrl":"https://doi.org/10.1080/10589759.2023.2273999","url":null,"abstract":"ABSTRACTTo take advantage of ultrasonic-based non-destructive testing (NDT) and data-driven intelligent defect diagnosis, the current study proposes a feature tensor classifier based on multi-source ultrasonic fusion, to enhance the defect diagnosis adaptability and reliability for gas-insulated switchgear (GIS) basin insulator. First, multi-source ultrasonic signals are acquired by finite element modelling (FEM), describing the healthy states of the GIS basin insulator completely. Second, time of flight (Tof)-featured tensors are expressed by wavelet transform (WT), and used to create the datasets. Third, a deep learning-based feature tensor classifier is proposed, and concerned training, validation, and testing processes are carried out. Finally, the effectiveness of feature tensor extraction is evaluated, and the anti-noise performance of the Tof-featured tensor classifier is verified. The main contributions indicate that the Tof-featured tensor classifier can realise excellent diagnosis performance, the average accuracy is, respectively, 90.53%, 99.75%, and 100% in training, validation, and testing sets, while the signal tensor classifier has poor performance. In addition, three other noised datasets are applied, and the result shows that the anti-noise performance of the Tof-featured tensor classifier is feasible, when SNR is greater than 1 dB.KEYWORDS: GIS basin insulatorintelligent fault diagnosismulti-source ultrasonic fusiontof-featured tensorconvolution neural network AcknowledgmentsMy heartfelt thanks are due to Prof. Han for his academic supervision and personal support. This work was supported by the Scientific and Technological Innovation Programs of Higher Education Institutions in Shanxi (Grant No. STIP2020L0699), Fund program of Key Laboratory of Signal Capturing & Processing, North University in Shanxi (Grant No. ISPT2020-8).Disclosure statementNo potential conflict of interest was reported by the author(s).","PeriodicalId":49746,"journal":{"name":"Nondestructive Testing and Evaluation","volume":"83 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-11-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135341562","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
ABSTRACTBecause oil and gas pipelines reside in high-temperature, high-pressure and highly acidic environments for a long time, it is very common for such pipes to experience corrosion thinning. To solve the long-term online monitoring problem of piezoelectric ultrasonic dry coupling of monitored objects with different surface roughnesses, the effects of different thicknesses of silver plates, different surface roughnesses and different coaxial loads on piezoelectric ultrasonic dry coupling were studied here. To improve the low SNR of the acquired signal, a Batworth high-pass filter and zero-phase digital filter are proposed for online ultrasonic monitoring. The effects of three kinds of roughness and axial load on the piezoelectric ultrasonic dry coupling signal are studied, and the relationship between the roughness, thickness of the silver plate and axial load is obtained. The results show that under different roughness conditions, the SNR of the ultrasonic signal increases with increasing axial load. When the axial load reaches a certain magnitude, the SNR of the ultrasonic signal tends to be stable. When the surface roughness is Ra = 1.6 µm and the thickness of the silver plate is 0.05 mm, the SNR of the ultrasonic signal reaches the best value, and the required axial load is minimal.KEYWORDS: Dry couplingpiezoelectric ultrasonicwall thickness monitoringsilver sheetroughness Disclosure statementNo potential conflict of interest was reported by the author(s).Additional informationFundingThis research was supported by the CNPC Major Science and Technology Project “Research on Key Equipment and Supporting Technology for Onshore Well Control Emergency Response” (2021ZZ03-1), the key project of CNPC “Development of 140MPa blowout Preventer (2021ZG08), and The Sichuan Science and Technology Plan Project (2022YFS0524).
{"title":"Influence of specimen roughness on dry coupling in piezoelectric ultrasonics","authors":"Shiqiang Wang, Laibin Zhang, Qiang Xu, Jianchun Fan, Jianbo Wu","doi":"10.1080/10589759.2023.2274131","DOIUrl":"https://doi.org/10.1080/10589759.2023.2274131","url":null,"abstract":"ABSTRACTBecause oil and gas pipelines reside in high-temperature, high-pressure and highly acidic environments for a long time, it is very common for such pipes to experience corrosion thinning. To solve the long-term online monitoring problem of piezoelectric ultrasonic dry coupling of monitored objects with different surface roughnesses, the effects of different thicknesses of silver plates, different surface roughnesses and different coaxial loads on piezoelectric ultrasonic dry coupling were studied here. To improve the low SNR of the acquired signal, a Batworth high-pass filter and zero-phase digital filter are proposed for online ultrasonic monitoring. The effects of three kinds of roughness and axial load on the piezoelectric ultrasonic dry coupling signal are studied, and the relationship between the roughness, thickness of the silver plate and axial load is obtained. The results show that under different roughness conditions, the SNR of the ultrasonic signal increases with increasing axial load. When the axial load reaches a certain magnitude, the SNR of the ultrasonic signal tends to be stable. When the surface roughness is Ra = 1.6 µm and the thickness of the silver plate is 0.05 mm, the SNR of the ultrasonic signal reaches the best value, and the required axial load is minimal.KEYWORDS: Dry couplingpiezoelectric ultrasonicwall thickness monitoringsilver sheetroughness Disclosure statementNo potential conflict of interest was reported by the author(s).Additional informationFundingThis research was supported by the CNPC Major Science and Technology Project “Research on Key Equipment and Supporting Technology for Onshore Well Control Emergency Response” (2021ZZ03-1), the key project of CNPC “Development of 140MPa blowout Preventer (2021ZG08), and The Sichuan Science and Technology Plan Project (2022YFS0524).","PeriodicalId":49746,"journal":{"name":"Nondestructive Testing and Evaluation","volume":"319 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-11-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135475060","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-11-06DOI: 10.1080/10589759.2023.2274008
Lei Wang, Zijie Chen, Hailin Zou, Dongsheng Huang, Yuanyuan Pan, Chak-Fong Cheang, Jianqing Li
ABSTRACTProlonged heat exposure may cause various physiological responses to outdoor workers. This will result in economic and productivity losses for a company and also may affect the sustainable development speed of cities. To avoid the above adverse effects, an alerting system is designed for outdoor workers to prevent them from overtime working in high-temperature scenarios. In the system, multiple sensors embedded micro-electromechanical system (MEMS) wearable device is used for real-time working status data collection, and a hybrid deep learning model is adopted to recognise the working status of outdoor workers. This hybrid model, called C-LSTM, combines the advantages of convolutional neural networks (CNN) and long short-term memory networks (LSTM) to extract useful spatial and temporal features of working status efficiently. Experimental results show that the performance on the inference time and accuracy of the C-LSTM model is better than that of conventional ones. The working status recognition accuracy of the C-LSTM model reaches 97.91%, and the inference time of the model reduces to less than 51 ms. In addition, the C-LSTM model has the best stability. The designed system can not only be used in outdoor high-temperature environment but also applied to medical and industrial scenarios, which further extends the application fields.KEYWORDS: Working statussensordeep learningsustainable smart city AcknowledgmentsThis research was funded in part by the Science and Technology Development Fund, Macao SAR under Grant No. 0047/2021/A, and in part by the National Social Science Fund of China under Grant No. 20BMZ053. We are also grateful for providing data by Shenzhen Topevery Technology Co., Ltd., Guangdong, China.Disclosure statementNo potential conflict of interest was reported by the author(s).Additional informationFundingThis work was supported by the Science and Technology Development Fund, Macao SAR under Grant [No. 0047/2021/A]; The National Social Science Fund of China under Grant [No. 20BMZ053].
{"title":"A deep learning-based high-temperature overtime working alert system for smart cities with multi-sensor data","authors":"Lei Wang, Zijie Chen, Hailin Zou, Dongsheng Huang, Yuanyuan Pan, Chak-Fong Cheang, Jianqing Li","doi":"10.1080/10589759.2023.2274008","DOIUrl":"https://doi.org/10.1080/10589759.2023.2274008","url":null,"abstract":"ABSTRACTProlonged heat exposure may cause various physiological responses to outdoor workers. This will result in economic and productivity losses for a company and also may affect the sustainable development speed of cities. To avoid the above adverse effects, an alerting system is designed for outdoor workers to prevent them from overtime working in high-temperature scenarios. In the system, multiple sensors embedded micro-electromechanical system (MEMS) wearable device is used for real-time working status data collection, and a hybrid deep learning model is adopted to recognise the working status of outdoor workers. This hybrid model, called C-LSTM, combines the advantages of convolutional neural networks (CNN) and long short-term memory networks (LSTM) to extract useful spatial and temporal features of working status efficiently. Experimental results show that the performance on the inference time and accuracy of the C-LSTM model is better than that of conventional ones. The working status recognition accuracy of the C-LSTM model reaches 97.91%, and the inference time of the model reduces to less than 51 ms. In addition, the C-LSTM model has the best stability. The designed system can not only be used in outdoor high-temperature environment but also applied to medical and industrial scenarios, which further extends the application fields.KEYWORDS: Working statussensordeep learningsustainable smart city AcknowledgmentsThis research was funded in part by the Science and Technology Development Fund, Macao SAR under Grant No. 0047/2021/A, and in part by the National Social Science Fund of China under Grant No. 20BMZ053. We are also grateful for providing data by Shenzhen Topevery Technology Co., Ltd., Guangdong, China.Disclosure statementNo potential conflict of interest was reported by the author(s).Additional informationFundingThis work was supported by the Science and Technology Development Fund, Macao SAR under Grant [No. 0047/2021/A]; The National Social Science Fund of China under Grant [No. 20BMZ053].","PeriodicalId":49746,"journal":{"name":"Nondestructive Testing and Evaluation","volume":"304 2","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-11-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135679084","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
ABSTRACTWith the rapid development of information technology, crowd-sourcing technology is increasingly used in non-invasive monitoring in smart cities. Applying recommendation algorithms in crowd-sourcing can optimise resource allocation, improve task-matching accuracy and enhance participant satisfaction, whereas existing recommendation algorithms cannot be directly applied in crowd-sourcing, as such scenarios have unique features, such as task timeliness and multi-role users. Designed explicitly for crowd-sourcing scenarios, our OPCA-CF (Outer-product Co-attention Collaborative Filtering) algorithm is formed by an upgraded ItemCF (Item-based Collaborative Filtering) algorithm as main-network and OPCA (Outer-product Co-attention) mechanism as a sub-network. Firstly, ItemCF is improved through attribute-level task feature learning, new-role feature and weighted cross-entropy in the loss function. Most importantly, we propose OPCA using outer-product, while the existing co-attention mechanism only uses inner-product. Compared with the best existing algorithm using real-world datasets, OPCA-CF’s performance is proved to be superior by 1.24%, 4.25% and 5.35%, with binary classification indicators AUC (Area under Curve), recommended Lists related indicators HR (Hit Ratio) and MRR (Mean Reciprocal Rank), respectively. All the performance indicators verified the effectiveness of the OPCA-CF algorithm.KEYWORDS: Recommendation algorithmattention mechanismcrowd-sourcingcollaborative filtering Disclosure statementNo potential conflict of interest was reported by the author(s).Additional informationFundingThis work was supported in part by the National Natural Science Foundation of China under Grant No. 62072319; the Sichuan Science and Technology Program under Grant No. 2023YFQ0022, 2022YFG0041, 2022YFG0155 and 2022YFG0157; the Luzhou Science and Technology Innovation R&D Program under Grant No. 2022CDLZ-6.
{"title":"A crowd-sourcing recommendation algorithm OPCA-CF using outer-product co-attention mechanism","authors":"Kejun Bi, Jingwen Liu, Qiwen Zhao, Yanru Chen, Bin Xing, Bing Guo","doi":"10.1080/10589759.2023.2273525","DOIUrl":"https://doi.org/10.1080/10589759.2023.2273525","url":null,"abstract":"ABSTRACTWith the rapid development of information technology, crowd-sourcing technology is increasingly used in non-invasive monitoring in smart cities. Applying recommendation algorithms in crowd-sourcing can optimise resource allocation, improve task-matching accuracy and enhance participant satisfaction, whereas existing recommendation algorithms cannot be directly applied in crowd-sourcing, as such scenarios have unique features, such as task timeliness and multi-role users. Designed explicitly for crowd-sourcing scenarios, our OPCA-CF (Outer-product Co-attention Collaborative Filtering) algorithm is formed by an upgraded ItemCF (Item-based Collaborative Filtering) algorithm as main-network and OPCA (Outer-product Co-attention) mechanism as a sub-network. Firstly, ItemCF is improved through attribute-level task feature learning, new-role feature and weighted cross-entropy in the loss function. Most importantly, we propose OPCA using outer-product, while the existing co-attention mechanism only uses inner-product. Compared with the best existing algorithm using real-world datasets, OPCA-CF’s performance is proved to be superior by 1.24%, 4.25% and 5.35%, with binary classification indicators AUC (Area under Curve), recommended Lists related indicators HR (Hit Ratio) and MRR (Mean Reciprocal Rank), respectively. All the performance indicators verified the effectiveness of the OPCA-CF algorithm.KEYWORDS: Recommendation algorithmattention mechanismcrowd-sourcingcollaborative filtering Disclosure statementNo potential conflict of interest was reported by the author(s).Additional informationFundingThis work was supported in part by the National Natural Science Foundation of China under Grant No. 62072319; the Sichuan Science and Technology Program under Grant No. 2023YFQ0022, 2022YFG0041, 2022YFG0155 and 2022YFG0157; the Luzhou Science and Technology Innovation R&D Program under Grant No. 2022CDLZ-6.","PeriodicalId":49746,"journal":{"name":"Nondestructive Testing and Evaluation","volume":"32 9","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-11-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135820376","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-11-02DOI: 10.1080/10589759.2023.2274000
Shihao Cui, Nan Wu, Pooneh Maghoul
ABSTRACTFatigue cracks, especially at their initial stage, can lead to a repetitive crack open-close breathing-like phenomenon in the vibration response of structural elements. As such, regularities, bi-linearity, or perturbations in the vibration response can arise. Entropy can be used to quantify the irregularity or bi-linearity in these responses since there is an apparent variation of entropy values on the two sides of a breathing crack. Here, we present a new breathing crack localisation method based on a spatially distributed entropy approach coupled with the empirical mode decomposition technique. To enhance the robustness, a pre-selection mechanism is proposed to select the most suitable entropy method. The proposed method is then employed to localise the breathing crack in a beam in a laboratory setup. It is concluded that the proposed approach can be effectively used for breathing crack localisation in a structural element.KEYWORDS: Crack localisationstructural health monitoringentropyempirical mode decompositionbreathing phenomenon Disclosure statementNo potential conflict of interest was reported by the author(s).Additional informationFundingThe first author is funded by the China Scholarship Council (CSC) from the Ministry of Education of P.R. China. This research was undertaken, in part, thanks to funding support from the Natural Sciences and Engineering Research Council of Canada (NSERC).
{"title":"Fatigue crack localisation based on empirical mode decomposition and pre-selected entropy","authors":"Shihao Cui, Nan Wu, Pooneh Maghoul","doi":"10.1080/10589759.2023.2274000","DOIUrl":"https://doi.org/10.1080/10589759.2023.2274000","url":null,"abstract":"ABSTRACTFatigue cracks, especially at their initial stage, can lead to a repetitive crack open-close breathing-like phenomenon in the vibration response of structural elements. As such, regularities, bi-linearity, or perturbations in the vibration response can arise. Entropy can be used to quantify the irregularity or bi-linearity in these responses since there is an apparent variation of entropy values on the two sides of a breathing crack. Here, we present a new breathing crack localisation method based on a spatially distributed entropy approach coupled with the empirical mode decomposition technique. To enhance the robustness, a pre-selection mechanism is proposed to select the most suitable entropy method. The proposed method is then employed to localise the breathing crack in a beam in a laboratory setup. It is concluded that the proposed approach can be effectively used for breathing crack localisation in a structural element.KEYWORDS: Crack localisationstructural health monitoringentropyempirical mode decompositionbreathing phenomenon Disclosure statementNo potential conflict of interest was reported by the author(s).Additional informationFundingThe first author is funded by the China Scholarship Council (CSC) from the Ministry of Education of P.R. China. This research was undertaken, in part, thanks to funding support from the Natural Sciences and Engineering Research Council of Canada (NSERC).","PeriodicalId":49746,"journal":{"name":"Nondestructive Testing and Evaluation","volume":"19 11","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-11-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135934389","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-11-02DOI: 10.1080/10589759.2023.2274011
Miao Yao, Yijing Lu, Jinteng Mou, Chen Yan, Dongjingdian Liu
ABSTRACTIn the smart city context, efficient urban surveillance under low-light conditions is crucial. Accurate object detection in dimly lit areas is vital for safety and nighttime driving. However, subpar, poorly lit images due to environmental or equipment limitations pose a challenge, affecting precision in tasks like object detection and segmentation. Existing solutions often involve time-consuming, inefficient image preprocessing and lack strong theoretical support for low-light city image enhancement. To address these issue, we propose an end-to-end pipeline named LAR-YOLO that leverages convolutional network to extract a set of image transformation parameters, and implements the Retinex theory to proficiently elevate the quality of the image. Unlike conventional approaches, this innovative method eliminates the need for hand-crafted parameters and can adaptively enhance each low-light image. Additionally, due to a restricted quantity of training data, the detection model may not achieve an adequate level of expertise to enhance detection accuracy. To tackle this challenge, we introduce a cross-domain learning approach that supplements the low-light model with knowledge from normal light scenarios. Our proof-of-principle experiments and ablation studies utilising ExDark and VOC datasets demonstrate that our proposed method outperforms similar low-light object detection algorithms by approximately 13% in terms of accuracy.KEYWORDS: Object detectionsmart cityRetinex theorylow-light image processingcross-domain learning AcknowledgmentsThis work was supported by the National Natural Science Foundation of China under Grant Nos. 62272462 and 51904294.Disclosure statementNo potential conflict of interest was reported by the author(s).Additional informationFundingThis work was supported by the National Natural Science Foundation of China [51904294]; National Natural Science Foundation of China [62272462].
{"title":"End-to-end adaptive object detection with learnable Retinex for low-light city environment","authors":"Miao Yao, Yijing Lu, Jinteng Mou, Chen Yan, Dongjingdian Liu","doi":"10.1080/10589759.2023.2274011","DOIUrl":"https://doi.org/10.1080/10589759.2023.2274011","url":null,"abstract":"ABSTRACTIn the smart city context, efficient urban surveillance under low-light conditions is crucial. Accurate object detection in dimly lit areas is vital for safety and nighttime driving. However, subpar, poorly lit images due to environmental or equipment limitations pose a challenge, affecting precision in tasks like object detection and segmentation. Existing solutions often involve time-consuming, inefficient image preprocessing and lack strong theoretical support for low-light city image enhancement. To address these issue, we propose an end-to-end pipeline named LAR-YOLO that leverages convolutional network to extract a set of image transformation parameters, and implements the Retinex theory to proficiently elevate the quality of the image. Unlike conventional approaches, this innovative method eliminates the need for hand-crafted parameters and can adaptively enhance each low-light image. Additionally, due to a restricted quantity of training data, the detection model may not achieve an adequate level of expertise to enhance detection accuracy. To tackle this challenge, we introduce a cross-domain learning approach that supplements the low-light model with knowledge from normal light scenarios. Our proof-of-principle experiments and ablation studies utilising ExDark and VOC datasets demonstrate that our proposed method outperforms similar low-light object detection algorithms by approximately 13% in terms of accuracy.KEYWORDS: Object detectionsmart cityRetinex theorylow-light image processingcross-domain learning AcknowledgmentsThis work was supported by the National Natural Science Foundation of China under Grant Nos. 62272462 and 51904294.Disclosure statementNo potential conflict of interest was reported by the author(s).Additional informationFundingThis work was supported by the National Natural Science Foundation of China [51904294]; National Natural Science Foundation of China [62272462].","PeriodicalId":49746,"journal":{"name":"Nondestructive Testing and Evaluation","volume":"10 6","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-11-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135936134","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-11-01DOI: 10.1080/10589759.2023.2274020
B Nidheesh Kumar, M C Santhosh Kumar, A Mercy Latha, Sachinlal Aroliveetil, M Nallaperumal, Krishnan Balasubramaniam, S Remakanthan, K K Moideenkutty, Shyam S Nair, L Mohan Kumar
ABSTRACTPolychlorotrifluoroethylene is used as a coating material over metallic seals in low-temperature applications to arrest fluid leakage from the impeller side in turbopumps. Typically, polychlorotrifluoroethylene coating is applied on V-type seals, with a thickness ranging from 80 to 130 μm by spraying an emulsion over the substrate followed by heat treatment. An attempt has been made to measure the polychlorotrifluoroethylene coating thickness over V-type seals using terahertz time-domain spectroscopy in reflection geometry, a noncontact, non-invasive NDT method. When the terahertz pulse from a transmitter photo-conductive antenna is incident on the V-type seal, it penetrates through the polychlorotrifluoroethylene coating. It gets reflected from the coating/base coat interface. Here, the reflected echoes from the air-to-polychlorotrifluoroethylene coating interface and polychlorotrifluoroethylene coating to the basecoat interface get overlapped in the time domain as the polychlorotrifluoroethylene coating layer is very thin. The sparse deconvolution technique separates the individual reflected signals and obtains the time delay signals from various interfaces. From the estimation of time delay values, the thickness of the coating has been computed using the refractive index value extracted using terahertz time-domain spectroscopy in transmission mode before the reflection measurements. The obtained thickness values are in close agreement with the coating thickness measured using optical microscopy.KEYWORDS: PolychlorotrifluroethyleneThz time-domain spectroscopythickness estimationreflection geometryseal AcknowledgmentsWe would like to thank Shri. Srirangam Siripothu and their team at PCM/VSSC for the support offered in sample preparation.Disclosure statementNo potential conflict of interest was reported by the author(s).Data availability statementThe data supporting this study’s findings are available from the corresponding authors upon reasonable request. No third-party data has been used for this research work.Additional informationFundingThe author(s) reported that there is no funding associated with the work featured in this article.
{"title":"Thickness measurement of polychlorotrifluoroethylene coating over metallic seal using terahertz time-domain spectroscopy","authors":"B Nidheesh Kumar, M C Santhosh Kumar, A Mercy Latha, Sachinlal Aroliveetil, M Nallaperumal, Krishnan Balasubramaniam, S Remakanthan, K K Moideenkutty, Shyam S Nair, L Mohan Kumar","doi":"10.1080/10589759.2023.2274020","DOIUrl":"https://doi.org/10.1080/10589759.2023.2274020","url":null,"abstract":"ABSTRACTPolychlorotrifluoroethylene is used as a coating material over metallic seals in low-temperature applications to arrest fluid leakage from the impeller side in turbopumps. Typically, polychlorotrifluoroethylene coating is applied on V-type seals, with a thickness ranging from 80 to 130 μm by spraying an emulsion over the substrate followed by heat treatment. An attempt has been made to measure the polychlorotrifluoroethylene coating thickness over V-type seals using terahertz time-domain spectroscopy in reflection geometry, a noncontact, non-invasive NDT method. When the terahertz pulse from a transmitter photo-conductive antenna is incident on the V-type seal, it penetrates through the polychlorotrifluoroethylene coating. It gets reflected from the coating/base coat interface. Here, the reflected echoes from the air-to-polychlorotrifluoroethylene coating interface and polychlorotrifluoroethylene coating to the basecoat interface get overlapped in the time domain as the polychlorotrifluoroethylene coating layer is very thin. The sparse deconvolution technique separates the individual reflected signals and obtains the time delay signals from various interfaces. From the estimation of time delay values, the thickness of the coating has been computed using the refractive index value extracted using terahertz time-domain spectroscopy in transmission mode before the reflection measurements. The obtained thickness values are in close agreement with the coating thickness measured using optical microscopy.KEYWORDS: PolychlorotrifluroethyleneThz time-domain spectroscopythickness estimationreflection geometryseal AcknowledgmentsWe would like to thank Shri. Srirangam Siripothu and their team at PCM/VSSC for the support offered in sample preparation.Disclosure statementNo potential conflict of interest was reported by the author(s).Data availability statementThe data supporting this study’s findings are available from the corresponding authors upon reasonable request. No third-party data has been used for this research work.Additional informationFundingThe author(s) reported that there is no funding associated with the work featured in this article.","PeriodicalId":49746,"journal":{"name":"Nondestructive Testing and Evaluation","volume":"62 3-4","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135272149","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-11-01DOI: 10.1080/10589759.2023.2274019
Xin Yang, Shuchang Zhang, Jiang Xu
ABSTRACTDebonding regions can occur at the interface of aluminium/rigid polyurethane foam composite plates (ARCP) during manufacturing. To help improve the technological process of manufacture, it is essential to precisely locate these debonding regions. This paper proposed a method to image the debonding regions in the ARCP based on the A0 mode Lamb waves. To study the influence of overlap between the debonding region and the coil of the electromagnetic acoustic transducer (EMAT) on Lamb wave propagation, a three-dimensional finite element simulation model was developed. A positive linear relationship was obtained between the amplitude of the A0 mode and the overlap rate between the debonding region and the coil of the EMAT, which was verified through experiments. Based on such relationship, an imaging method was proposed. The imaging method consisted a horizontal scanning for the entire sample and vertical scanning for possible debonding regions determined from the horizontal scanning. The result of the imaging experiments can precisely reveal the size and position of the debonding regions, the maximum relative error of the centre position of the defect is 7.5% and the maximum relative deviation of the dimensions is 16.0%. This imaging method can serve as a reference for debonding imaging in composite plates.KEYWORDS: Aluminium/rigid polyurethane foam composite platesdebonding imagingelectromagnetic acoustic transducersA0 mode Lamb waves Disclosure statementNo potential conflict of interest was reported by the author(s).
{"title":"Debonding imaging of the aluminium/rigid polyurethane foam composite plates using A <sub>0</sub> mode Lamb waves","authors":"Xin Yang, Shuchang Zhang, Jiang Xu","doi":"10.1080/10589759.2023.2274019","DOIUrl":"https://doi.org/10.1080/10589759.2023.2274019","url":null,"abstract":"ABSTRACTDebonding regions can occur at the interface of aluminium/rigid polyurethane foam composite plates (ARCP) during manufacturing. To help improve the technological process of manufacture, it is essential to precisely locate these debonding regions. This paper proposed a method to image the debonding regions in the ARCP based on the A0 mode Lamb waves. To study the influence of overlap between the debonding region and the coil of the electromagnetic acoustic transducer (EMAT) on Lamb wave propagation, a three-dimensional finite element simulation model was developed. A positive linear relationship was obtained between the amplitude of the A0 mode and the overlap rate between the debonding region and the coil of the EMAT, which was verified through experiments. Based on such relationship, an imaging method was proposed. The imaging method consisted a horizontal scanning for the entire sample and vertical scanning for possible debonding regions determined from the horizontal scanning. The result of the imaging experiments can precisely reveal the size and position of the debonding regions, the maximum relative error of the centre position of the defect is 7.5% and the maximum relative deviation of the dimensions is 16.0%. This imaging method can serve as a reference for debonding imaging in composite plates.KEYWORDS: Aluminium/rigid polyurethane foam composite platesdebonding imagingelectromagnetic acoustic transducersA0 mode Lamb waves Disclosure statementNo potential conflict of interest was reported by the author(s).","PeriodicalId":49746,"journal":{"name":"Nondestructive Testing and Evaluation","volume":"315 2","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135321694","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Crack localisation in composite cantilever beams using natural frequency measurements","authors":"Mustapha Dahak, Noureddine Touat, Tarak Benkedjouh","doi":"10.1080/10589759.2023.2274017","DOIUrl":"https://doi.org/10.1080/10589759.2023.2274017","url":null,"abstract":"","PeriodicalId":49746,"journal":{"name":"Nondestructive Testing and Evaluation","volume":"49 2","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-10-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"136103582","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：