PCB Soldering Defect Inspection Using Multitask Learning under Low Data Regimes
To increase the reliability of the printed circuit board (PCB) manufacturing process, automated optical inspection is often employed for soldering defect detection. However, traditional approaches built on handcrafted features, predefined rules, or thresholds are often susceptible to the variation o...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Wiley
2023-12-01
|
| Series: | Advanced Intelligent Systems |
| Subjects: | |
| Online Access: | https://doi.org/10.1002/aisy.202300364 |
| _version_ | 1797378782595645440 |
|---|---|
| author | Sik-Ho Tsang Zhaoqing Suo Tom Tak-Lam Chan Huu-Thanh Nguyen Daniel Pak-Kong Lun |
| author_facet | Sik-Ho Tsang Zhaoqing Suo Tom Tak-Lam Chan Huu-Thanh Nguyen Daniel Pak-Kong Lun |
| author_sort | Sik-Ho Tsang |
| collection | DOAJ |
| description | To increase the reliability of the printed circuit board (PCB) manufacturing process, automated optical inspection is often employed for soldering defect detection. However, traditional approaches built on handcrafted features, predefined rules, or thresholds are often susceptible to the variation of the acquired images’ quality and give unstable performances. To solve this problem, a deep learning‐based soldering defect detection method is developed in this article. Like many real‐life deep learning applications, the number of available training samples is often limited. This creates a challenging low‐data scenario, as deep learning typically requires massive data to perform well. To address this issue, a multitask learning model is proposed, namely, PCBMTL, that can simultaneously learn the classification and segmentation tasks under low‐data regimes. By acquiring the segmentation knowledge, classification performance is substantially improved with few samples. To facilitate the study, a soldering defect image dataset, namely, PCBSPDefect, is built. It focuses on the dual in‐line packages (DIP) at the PCB back side, DIP at the PCB front side, and flat flexible cables. Experimental results show that the proposed PCBMTL outperforms the best existing approaches by over 5–17% of average accuracy for different datasets. |
| first_indexed | 2024-03-08T20:12:36Z |
| format | Article |
| id | doaj.art-ec980cde730b435f808567ccf9a67f56 |
| institution | Directory Open Access Journal |
| issn | 2640-4567 |
| language | English |
| last_indexed | 2024-03-08T20:12:36Z |
| publishDate | 2023-12-01 |
| publisher | Wiley |
| record_format | Article |
| series | Advanced Intelligent Systems |
| spelling | doaj.art-ec980cde730b435f808567ccf9a67f562023-12-23T04:53:50ZengWileyAdvanced Intelligent Systems2640-45672023-12-01512n/an/a10.1002/aisy.202300364PCB Soldering Defect Inspection Using Multitask Learning under Low Data RegimesSik-Ho Tsang0Zhaoqing Suo1Tom Tak-Lam Chan2Huu-Thanh Nguyen3Daniel Pak-Kong Lun4Unit 1212-1213 Hong Kong Science Park 12/F, Building 19W, Pak Shek Kok, NT Hong Kong ChinaUnit 1212-1213 Hong Kong Science Park 12/F, Building 19W, Pak Shek Kok, NT Hong Kong ChinaUnit 1212-1213 Hong Kong Science Park 12/F, Building 19W, Pak Shek Kok, NT Hong Kong ChinaUnit 1212-1213 Hong Kong Science Park 12/F, Building 19W, Pak Shek Kok, NT Hong Kong ChinaUnit 1212-1213 Hong Kong Science Park 12/F, Building 19W, Pak Shek Kok, NT Hong Kong ChinaTo increase the reliability of the printed circuit board (PCB) manufacturing process, automated optical inspection is often employed for soldering defect detection. However, traditional approaches built on handcrafted features, predefined rules, or thresholds are often susceptible to the variation of the acquired images’ quality and give unstable performances. To solve this problem, a deep learning‐based soldering defect detection method is developed in this article. Like many real‐life deep learning applications, the number of available training samples is often limited. This creates a challenging low‐data scenario, as deep learning typically requires massive data to perform well. To address this issue, a multitask learning model is proposed, namely, PCBMTL, that can simultaneously learn the classification and segmentation tasks under low‐data regimes. By acquiring the segmentation knowledge, classification performance is substantially improved with few samples. To facilitate the study, a soldering defect image dataset, namely, PCBSPDefect, is built. It focuses on the dual in‐line packages (DIP) at the PCB back side, DIP at the PCB front side, and flat flexible cables. Experimental results show that the proposed PCBMTL outperforms the best existing approaches by over 5–17% of average accuracy for different datasets.https://doi.org/10.1002/aisy.202300364automated optical inspectiondeep neural networksmultitask learningprinted circuit boardssoldering defect detection |
| spellingShingle | Sik-Ho Tsang Zhaoqing Suo Tom Tak-Lam Chan Huu-Thanh Nguyen Daniel Pak-Kong Lun PCB Soldering Defect Inspection Using Multitask Learning under Low Data Regimes Advanced Intelligent Systems automated optical inspection deep neural networks multitask learning printed circuit boards soldering defect detection |
| title | PCB Soldering Defect Inspection Using Multitask Learning under Low Data Regimes |
| title_full | PCB Soldering Defect Inspection Using Multitask Learning under Low Data Regimes |
| title_fullStr | PCB Soldering Defect Inspection Using Multitask Learning under Low Data Regimes |
| title_full_unstemmed | PCB Soldering Defect Inspection Using Multitask Learning under Low Data Regimes |
| title_short | PCB Soldering Defect Inspection Using Multitask Learning under Low Data Regimes |
| title_sort | pcb soldering defect inspection using multitask learning under low data regimes |
| topic | automated optical inspection deep neural networks multitask learning printed circuit boards soldering defect detection |
| url | https://doi.org/10.1002/aisy.202300364 |
| work_keys_str_mv | AT sikhotsang pcbsolderingdefectinspectionusingmultitasklearningunderlowdataregimes AT zhaoqingsuo pcbsolderingdefectinspectionusingmultitasklearningunderlowdataregimes AT tomtaklamchan pcbsolderingdefectinspectionusingmultitasklearningunderlowdataregimes AT huuthanhnguyen pcbsolderingdefectinspectionusingmultitasklearningunderlowdataregimes AT danielpakkonglun pcbsolderingdefectinspectionusingmultitasklearningunderlowdataregimes |