Anomaly Detection for Solder Joints Using β-VAE

In the assembly process of printed circuit boards (PCBs), most of the errors are caused by solder joints in surface mount devices (SMDs). In the literature, traditional feature extraction-based methods require designing hand-crafted features and rely on the tiered red green blue (RGB) illumination t...

Full description

Saved in:

Bibliographic Details
Published in:	IEEE transactions on components, packaging, and manufacturing technology (2011) Vol. 11; no. 12; pp. 2214 - 2221
Main Authors:	Ulger, Furkan, Yuksel, Seniha Esen, Yilmaz, Atila
Format:	Journal Article
Language:	English
Published:	Piscataway IEEE 01.12.2021 The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Subjects:	<italic xmlns:ali="http://www.niso.org/schemas/ali/1.0/" xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance">β -variational autoencoder (VAE) Accuracy Anomalies Anomaly detection Artificial neural networks automated optical inspection (AOI) Automatic optical inspection Circuit boards Errors Feature extraction Illumination Inspection Integrated circuits Model accuracy Representations solder joint inspection (SJI) Soldered joints Soldering Solders Surface-mounted devices Training unsupervised anomaly detection VAE
ISSN:	2156-3950, 2156-3985
Online Access:	Get full text
Tags:	Add Tag No Tags, Be the first to tag this record!

Abstract	In the assembly process of printed circuit boards (PCBs), most of the errors are caused by solder joints in surface mount devices (SMDs). In the literature, traditional feature extraction-based methods require designing hand-crafted features and rely on the tiered red green blue (RGB) illumination to detect solder joint errors, whereas the supervised convolutional neural network (CNN)-based approaches require a lot of labeled abnormal samples (defective solder joints) to achieve high accuracy. To solve the optical inspection problem in unrestricted environments with no special lighting and without the existence of error-free reference boards, we propose a new beta-variational autoencoder (beta-VAE) architecture for anomaly detection that can work on both integrated circuit (IC) and non-IC components. We show that the proposed model learns disentangled representation of data, leading to more independent features and improved latent space representations. We compare the activation and gradient-based representations that are used to characterize anomalies and observe the effect of different beta parameters on accuracy and untwining the feature representations in beta-VAE. Finally, we show that anomalies on solder joints can be detected with high accuracy via a model trained directly on normal samples without designated hardware or feature engineering.
AbstractList	In the assembly process of printed circuit boards (PCBs), most of the errors are caused by solder joints in surface mount devices (SMDs). In the literature, traditional feature extraction-based methods require designing hand-crafted features and rely on the tiered red green blue (RGB) illumination to detect solder joint errors, whereas the supervised convolutional neural network (CNN)-based approaches require a lot of labeled abnormal samples (defective solder joints) to achieve high accuracy. To solve the optical inspection problem in unrestricted environments with no special lighting and without the existence of error-free reference boards, we propose a new beta-variational autoencoder (beta-VAE) architecture for anomaly detection that can work on both integrated circuit (IC) and non-IC components. We show that the proposed model learns disentangled representation of data, leading to more independent features and improved latent space representations. We compare the activation and gradient-based representations that are used to characterize anomalies and observe the effect of different beta parameters on accuracy and untwining the feature representations in beta-VAE. Finally, we show that anomalies on solder joints can be detected with high accuracy via a model trained directly on normal samples without designated hardware or feature engineering.
Author	Ulger, Furkan Yuksel, Seniha Esen Yilmaz, Atila
Author_xml	– sequence: 1 givenname: Furkan orcidid: 0000-0002-5935-2743 surname: Ulger fullname: Ulger, Furkan email: furkan.ulger@stu.ee.hacettepe.edu.tr organization: Department of Electrical and Electronics Engineering, Hacettepe University, Ankara, Turkey – sequence: 2 givenname: Seniha Esen orcidid: 0000-0002-8868-1132 surname: Yuksel fullname: Yuksel, Seniha Esen email: eyuksel@ee.hacettepe.edu.tr organization: Department of Electrical and Electronics Engineering, Hacettepe University, Ankara, Turkey – sequence: 3 givenname: Atila surname: Yilmaz fullname: Yilmaz, Atila email: ayilmaz@ee.hacettepe.edu.tr organization: Department of Electrical and Electronics Engineering, Hacettepe University, Ankara, Turkey
BookMark	eNp9kL9OwzAQhy1UJErpC8ASiTnhzo7tZKxK-acikGhZLTexUao0LnY69LV4EJ6JlFYdGLjlbvh9d7rvnPQa1xhCLhESRMhvZuPX51lCgWLCkCIV_IT0KXIRszzjvePM4YwMQ1hCVzwDCaxPYNS4la630a1pTdFWroms89Gbq0vjoydXNW2I5qFqPqLvr_h9NLkgp1bXwQwPfUDmd5PZ-CGevtw_jkfTuKA5b2NWMKlxUepUC2FQC85oYQqpWYmZTXMLCCkHhsLK0pbGMrZITQmS2iwHsGxArvd71959bkxo1dJtfNOdVFQgoERJRZfK9qnCuxC8saqoWr17o_W6qhWC2ilSv4rUTpE6KOpQ-gdd-2ql_fZ_6GoPVcaYI5BzmaeUsR_denKX
CODEN	ITCPC8
CitedBy_id	crossref_primary_10_3390_e25020268 crossref_primary_10_1007_s00170_025_15383_4 crossref_primary_10_1109_TCPMT_2022_3224997 crossref_primary_10_3389_frobt_2025_1554196 crossref_primary_10_1109_TIM_2023_3277935 crossref_primary_10_1016_j_media_2024_103229 crossref_primary_10_1007_s00170_025_15460_8 crossref_primary_10_3390_asi7010011 crossref_primary_10_1088_1361_6501_ac769a crossref_primary_10_3390_s24030738 crossref_primary_10_1016_j_addma_2024_104635 crossref_primary_10_1177_09544070251319563 crossref_primary_10_3390_bioengineering10060683 crossref_primary_10_1007_s10796_024_10507_9 crossref_primary_10_1016_j_measurement_2025_117443 crossref_primary_10_1007_s10015_025_01028_y crossref_primary_10_1007_s10489_025_06449_7 crossref_primary_10_3389_fmtec_2024_1277152 crossref_primary_10_1016_j_measurement_2024_116392 crossref_primary_10_3390_machines12090603
Cites_doi	10.1109/TPAMI.2013.50 10.1109/TCPMT.2018.2789453 10.1109/TCPMT.2012.2231902 10.1007/s00170-006-0730-0 10.1126/science.1127647 10.1016/j.aei.2019.100933 10.1109/34.3902 10.23919/SPA.2019.8936659 10.1109/6104.846932 10.1007/978-3-030-32251-9_32 10.1109/TCPMT.2011.2168531 10.1016/j.rcim.2014.03.003 10.1007/s00170-018-3022-6 10.1016/j.aei.2019.101004 10.5772/51699 10.1006/cviu.1996.0020 10.1016/S0031-3203(98)00103-4 10.1117/1.JEI.29.4.041013 10.1109/TCPMT.2011.2118208 10.1016/j.rcim.2011.03.007 10.1007/978-3-030-58589-1_13 10.1109/TII.2006.877265 10.1145/3383261
ContentType	Journal Article
Copyright	Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2021
Copyright_xml	– notice: Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2021
DBID	97E RIA RIE AAYXX CITATION 7SP 8FD F28 FR3 L7M
DOI	10.1109/TCPMT.2021.3121265
DatabaseName	IEEE All-Society Periodicals Package (ASPP) 2005–Present IEEE All-Society Periodicals Package (ASPP) 1998–Present IEEE Electronic Library (IEL) CrossRef Electronics & Communications Abstracts Technology Research Database ANTE: Abstracts in New Technology & Engineering Engineering Research Database Advanced Technologies Database with Aerospace
DatabaseTitle	CrossRef Engineering Research Database Technology Research Database Advanced Technologies Database with Aerospace ANTE: Abstracts in New Technology & Engineering Electronics & Communications Abstracts
DatabaseTitleList	Engineering Research Database
Database_xml	– sequence: 1 dbid: RIE name: IEEE Electronic Library (IEL) url: https://ieeexplore.ieee.org/ sourceTypes: Publisher
DeliveryMethod	fulltext_linktorsrc
Discipline	Engineering
EISSN	2156-3985
EndPage	2221
ExternalDocumentID	10_1109_TCPMT_2021_3121265 9579423
Genre	orig-research
GrantInformation_xml	– fundername: Aselsan Inc., through a project between Hacettepe University and Aselsan Inc funderid: 10.13039/501100007276
GroupedDBID	0R~ 4.4 6IK 97E AAJGR AARMG AASAJ AAWTH ABAZT ABQJQ ABVLG ACIWK AENEX AGQYO AGSQL AHBIQ AKJIK AKQYR ALMA_UNASSIGNED_HOLDINGS ATWAV BEFXN BFFAM BGNUA BKEBE BPEOZ EBS EJD IFIPE IPLJI JAVBF M43 OCL RIA RIE RNS AAYXX CITATION 7SP 8FD F28 FR3 L7M
ID	FETCH-LOGICAL-c295t-3c37a1bda4a66e1a6532cec7a3d18f49f010450316f7dfdef33b4ed072f8900f3
IEDL.DBID	RIE
ISICitedReferencesCount	34
ISICitedReferencesURI	http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=000730529000024&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN	2156-3950
IngestDate	Mon Jun 30 10:07:33 EDT 2025 Tue Nov 18 22:38:30 EST 2025 Sat Nov 29 06:08:13 EST 2025 Wed Aug 27 05:07:52 EDT 2025
IsPeerReviewed	true
IsScholarly	true
Issue	12
Language	English
License	https://ieeexplore.ieee.org/Xplorehelp/downloads/license-information/IEEE.html https://doi.org/10.15223/policy-029 https://doi.org/10.15223/policy-037
LinkModel	DirectLink
MergedId	FETCHMERGED-LOGICAL-c295t-3c37a1bda4a66e1a6532cec7a3d18f49f010450316f7dfdef33b4ed072f8900f3
Notes	ObjectType-Article-1 SourceType-Scholarly Journals-1 ObjectType-Feature-2 content type line 14
ORCID	0000-0002-5935-2743 0000-0002-8868-1132
PQID	2610171726
PQPubID	1006342
PageCount	8
ParticipantIDs	crossref_citationtrail_10_1109_TCPMT_2021_3121265 crossref_primary_10_1109_TCPMT_2021_3121265 ieee_primary_9579423 proquest_journals_2610171726
PublicationCentury	2000
PublicationDate	2021-12-01
PublicationDateYYYYMMDD	2021-12-01
PublicationDate_xml	– month: 12 year: 2021 text: 2021-12-01 day: 01
PublicationDecade	2020
PublicationPlace	Piscataway
PublicationPlace_xml	– name: Piscataway
PublicationTitle	IEEE transactions on components, packaging, and manufacturing technology (2011)
PublicationTitleAbbrev	TCPMT
PublicationYear	2021
Publisher	IEEE The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Publisher_xml	– name: IEEE – name: The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
References	kingma (ref34) 2014 wu (ref10) 2008 ref12 ref15 zong (ref23) 2018 ref30 ref11 lu (ref31) 2020; 2020 ref2 kingma (ref4) 2013 ref1 ref17 ref16 ioffe (ref33) 2015 ref19 ref18 van der maaten (ref35) 2008; 9 ko (ref14) 2000; 23 ref24 lu (ref26) 2018 ref25 ref20 ref22 ref21 wu (ref13) 2011; 1 wada (ref32) 2016 ref28 ref27 ref8 rezende (ref5) 2014 ref7 boracchi (ref29) 2020 ref9 ref6 higgins (ref3) 2017
References_xml	– ident: ref6 doi: 10.1109/TPAMI.2013.50 – ident: ref21 doi: 10.1109/TCPMT.2018.2789453 – ident: ref9 doi: 10.1109/TCPMT.2012.2231902 – start-page: 1 year: 2018 ident: ref23 article-title: Deep autoencoding Gaussian mixture model for unsupervised anomaly detection publication-title: Proc ICLR – ident: ref15 doi: 10.1007/s00170-006-0730-0 – ident: ref30 doi: 10.1126/science.1127647 – year: 2013 ident: ref4 article-title: Auto-encoding variational Bayes publication-title: arXiv 1312 6114 – year: 2016 ident: ref32 publication-title: Labelme Image Polygonal Annotation with Python – ident: ref22 doi: 10.1016/j.aei.2019.100933 – year: 2018 ident: ref26 article-title: Anomaly detection for skin disease images using variational autoencoder publication-title: arXiv 1807 01349 – start-page: 1 year: 2017 ident: ref3 article-title: Beta-VAE: Learning basic visual concepts with a constrained variational framework publication-title: Proc ICLR – ident: ref19 doi: 10.1109/34.3902 – ident: ref17 doi: 10.23919/SPA.2019.8936659 – volume: 23 start-page: 93 year: 2000 ident: ref14 article-title: Solder joints inspection using a neural network and fuzzy rule-based classification method publication-title: IEEE Trans Electron Packag Manuf doi: 10.1109/6104.846932 – ident: ref27 doi: 10.1007/978-3-030-32251-9_32 – ident: ref8 doi: 10.1109/TCPMT.2011.2168531 – ident: ref11 doi: 10.1016/j.rcim.2014.03.003 – ident: ref12 doi: 10.1007/s00170-018-3022-6 – start-page: 91 year: 2020 ident: ref29 article-title: Tutorial: Anomaly detection in images publication-title: Proc IEEE ICIP – volume: 2020 start-page: 366 year: 2020 ident: ref31 article-title: FICS-PCB: A multi-modal image dataset for automated printed circuit board visual inspection publication-title: IACR Cryptol ePrint Arch – ident: ref20 doi: 10.1016/j.aei.2019.101004 – ident: ref2 doi: 10.5772/51699 – ident: ref1 doi: 10.1006/cviu.1996.0020 – ident: ref7 doi: 10.1016/S0031-3203(98)00103-4 – ident: ref24 doi: 10.1117/1.JEI.29.4.041013 – start-page: 240 year: 2008 ident: ref10 article-title: An AOI algorithm for PCB based on feature extraction publication-title: Proc 7th World Congr Intell Control Autom – start-page: 448 year: 2015 ident: ref33 article-title: Batch normalization: Accelerating deep network training by reducing internal covariate shift publication-title: Proc Int Conf Mach Learn – volume: 1 start-page: 689 year: 2011 ident: ref13 article-title: Feature-extraction-based inspection algorithm for IC solder joints publication-title: IEEE Trans Compon Packag Manuf Technol doi: 10.1109/TCPMT.2011.2118208 – ident: ref18 doi: 10.1016/j.rcim.2011.03.007 – ident: ref28 doi: 10.1007/978-3-030-58589-1_13 – start-page: 1278 year: 2014 ident: ref5 article-title: Stochastic backpropagation and approximate inference in deep generative models publication-title: Proc Int Conf Mach Learn – year: 2014 ident: ref34 article-title: Adam: A method for stochastic optimization publication-title: arXiv 1412 6980 – ident: ref16 doi: 10.1109/TII.2006.877265 – ident: ref25 doi: 10.1145/3383261 – volume: 9 start-page: 2580 year: 2008 ident: ref35 article-title: Visualizing data using t-SNE publication-title: J Mach Learn Res
SSID	ssj0000580703
Score	2.4375315
Snippet	In the assembly process of printed circuit boards (PCBs), most of the errors are caused by solder joints in surface mount devices (SMDs). In the literature,...
SourceID	proquest crossref ieee
SourceType	Aggregation Database Enrichment Source Index Database Publisher
StartPage	2214
SubjectTerms	<italic xmlns:ali="http://www.niso.org/schemas/ali/1.0/" xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance">β -variational autoencoder (VAE) Accuracy Anomalies Anomaly detection Artificial neural networks automated optical inspection (AOI) Automatic optical inspection Circuit boards Errors Feature extraction Illumination Inspection Integrated circuits Model accuracy Representations solder joint inspection (SJI) Soldered joints Soldering Solders Surface-mounted devices Training unsupervised anomaly detection VAE
Title	Anomaly Detection for Solder Joints Using β-VAE
URI	https://ieeexplore.ieee.org/document/9579423 https://www.proquest.com/docview/2610171726
Volume	11
WOSCitedRecordID	wos000730529000024&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
hasFullText	1
inHoldings	1
isFullTextHit
isPrint
journalDatabaseRights	– providerCode: PRVIEE databaseName: IEEE Electronic Library (IEL) customDbUrl: eissn: 2156-3985 dateEnd: 99991231 omitProxy: false ssIdentifier: ssj0000580703 issn: 2156-3950 databaseCode: RIE dateStart: 20110101 isFulltext: true titleUrlDefault: https://ieeexplore.ieee.org/ providerName: IEEE
link	http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV1dS8MwFL1M8UEf_JridEoffNO4tFma5nHMDREUwSl7K1l6A4J2slXBv-UP8TeZpF0ZKIJvpSSlnJubm5Pc3ANwat8y638RQY4J6XKFRAljSMiV5CyzMTujXmxC3N4m47G8a8B5fRcGEX3yGV64R3-Wn031m9sq67gjJRv-V2BFiLi8q1Xvp1CeuNHrtOQsJSFMcrq4I0NlZ9S_uxlZNhiFlqTa2drFkqU45IVVfszGPsQMt_73c9uwWS0lg15p-x1oYL4LG0sFBptALbt_Uc8fwSUWPucqD-wiNbifOm3u4Hr6lBfzwGcNBF-f5LE32IOH4WDUvyKVRgLRkeQFYZoJFU4y1VVxjKGKOYs0aqFYFiamK43jW9x6bmxEZjI0jE26mFERmURSatg-rObTHA8gQD1hSmkTRtZ8oWLKGMoMWg9nOk5i2YJwgViqqwLiTsfiOfVEgsrUo5w6lNMK5Rac1X1ey_IZf7ZuOlzrlhWkLWgvDJNWHjZPLfNzpX5EFB_-3usI1t23y9STNqwWszc8hjX9XjzNZyd-8HwDV_DANQ
linkProvider	IEEE
linkToHtml	http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV1dS8MwFL3oFNQHv8Xp1D74pnFp07TNo-iGH9sQnLK3kqU3MNBOXCf4t_wh_iaTrBuCIvhWSkLLubm5OcnNPQDH5i0z_hcQ5JiQkEskMtaa-FwKzjITszPqxCbiTifp9cTdHJzO7sIgoks-wzP76M7ys6Ea262yuj1SMuF_HhZ4GAZ0cltrtqNCeWLHr1WTM6SEMMHp9JYMFfXuxV27a_hg4BuaauZrG02-RSInrfJjPnZBprn2v99bh9VyMemdT6y_AXOYb8LKtxKDW0ANv3-WT-_eJRYu6yr3zDLVux9adW7vZjjIi5Hn8ga8zw_yeN7Yhodmo3txRUqVBKICwQvCFIul389kKKMIfRlxFihUsWSZn-hQaMu4uPHdSMeZzlAz1g8xo3GgE0GpZjtQyYc57oKHqs-kVNoPjAF9yaTWlGk0Ps5UlESiCv4UsVSVJcStksVT6qgEFalDObUopyXKVTiZ9XmZFND4s_WWxXXWsoS0CrWpYdLSx0ap4X622E8cRHu_9zqCpatuu5W2rju3-7BsvzNJRKlBpXgd4wEsqrdiMHo9dAPpC4YZw3w
openUrl	ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=Anomaly+Detection+for+Solder+Joints+Using+%CE%B2+-VAE&rft.jtitle=IEEE+transactions+on+components%2C+packaging%2C+and+manufacturing+technology+%282011%29&rft.au=Ulger%2C+Furkan&rft.au=Yuksel%2C+Seniha+Esen&rft.au=Yilmaz%2C+Atila&rft.date=2021-12-01&rft.issn=2156-3950&rft.eissn=2156-3985&rft.volume=11&rft.issue=12&rft.spage=2214&rft.epage=2221&rft_id=info:doi/10.1109%2FTCPMT.2021.3121265&rft.externalDBID=n%2Fa&rft.externalDocID=10_1109_TCPMT_2021_3121265
thumbnail_l	http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=2156-3950&client=summon
thumbnail_m	http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=2156-3950&client=summon
thumbnail_s	http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=2156-3950&client=summon