Deep-Learning Methods for Defect Inspection of Plated Through Holes With Clustering-Based Auto-Labeling and GAN-Based Model Training

This paper presents the integration of several deep learning techniques for defect inspection of plated through-hole (PTH) on printed circuit boards (PCBs). In our proposed system, the object detection technology of You Only Look Once (YOLO) allocates the position of PTHs; a semi-automatic clusterin...

Full description

Saved in:
Bibliographic Details
Main Authors: Chang-Yeh Hsieh, Ling-Shen Tseng, Yi-Han Chen, Chiung-Hui Tsai, Chih-Hung Wu
Format: Article
Language:English
Published: IEEE 2024-01-01
Series:IEEE Access
Subjects:
Online Access:https://ieeexplore.ieee.org/document/10792891/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850250095162818560
author Chang-Yeh Hsieh
Ling-Shen Tseng
Yi-Han Chen
Chiung-Hui Tsai
Chih-Hung Wu
author_facet Chang-Yeh Hsieh
Ling-Shen Tseng
Yi-Han Chen
Chiung-Hui Tsai
Chih-Hung Wu
author_sort Chang-Yeh Hsieh
collection DOAJ
description This paper presents the integration of several deep learning techniques for defect inspection of plated through-hole (PTH) on printed circuit boards (PCBs). In our proposed system, the object detection technology of You Only Look Once (YOLO) allocates the position of PTHs; a semi-automatic clustering mechanism distinguishes normal and defective PTHs for collecting training data. A convolution neural network based on the ResNet framework is established for detecting PTH defects. A retrain mechanism is designed for retracting misclassified PTHs and updating the recognition model. The generative adversarial network (GAN) is employed to deal with the insufficiency of negative training samples. The detection model can achieve 98.96% accuracy with continuous retraining and data augmentation. Further, a filtering mechanism based on template analysis eliminates ambiguous outputs of GAN and improves the accuracy of the detection model. This study integrates these deep learning techniques to establish a PTH defect detection system tested in a PCB manufactory against real PTHs and gained an overall prediction accuracy of 99.48%.
format Article
id doaj-art-e002fbe0ceb04cbf95b41523b5cbbc46
institution OA Journals
issn 2169-3536
language English
publishDate 2024-01-01
publisher IEEE
record_format Article
series IEEE Access
spelling doaj-art-e002fbe0ceb04cbf95b41523b5cbbc462025-08-20T01:58:19ZengIEEEIEEE Access2169-35362024-01-011219059819061010.1109/ACCESS.2024.351546310792891Deep-Learning Methods for Defect Inspection of Plated Through Holes With Clustering-Based Auto-Labeling and GAN-Based Model TrainingChang-Yeh Hsieh0Ling-Shen Tseng1https://orcid.org/0000-0002-6661-6815Yi-Han Chen2https://orcid.org/0000-0003-1054-0353Chiung-Hui Tsai3https://orcid.org/0000-0002-2250-4715Chih-Hung Wu4https://orcid.org/0000-0002-8109-068XDepartment of Electrical Engineering, National University of Kaohsiung, Kaohsiung, TaiwanDepartment of Electrical Engineering, National University of Kaohsiung, Kaohsiung, TaiwanDepartment of Electrical Engineering, National University of Kaohsiung, Kaohsiung, TaiwanDepartment of Electrical Engineering, National University of Kaohsiung, Kaohsiung, TaiwanDepartment of Electrical Engineering, National University of Kaohsiung, Kaohsiung, TaiwanThis paper presents the integration of several deep learning techniques for defect inspection of plated through-hole (PTH) on printed circuit boards (PCBs). In our proposed system, the object detection technology of You Only Look Once (YOLO) allocates the position of PTHs; a semi-automatic clustering mechanism distinguishes normal and defective PTHs for collecting training data. A convolution neural network based on the ResNet framework is established for detecting PTH defects. A retrain mechanism is designed for retracting misclassified PTHs and updating the recognition model. The generative adversarial network (GAN) is employed to deal with the insufficiency of negative training samples. The detection model can achieve 98.96% accuracy with continuous retraining and data augmentation. Further, a filtering mechanism based on template analysis eliminates ambiguous outputs of GAN and improves the accuracy of the detection model. This study integrates these deep learning techniques to establish a PTH defect detection system tested in a PCB manufactory against real PTHs and gained an overall prediction accuracy of 99.48%.https://ieeexplore.ieee.org/document/10792891/Automated optical inspection (AOI)defect detectiondeep learningelectroplated through-hole (PTH)printed circuit board (PCB)generative adversarial network (GAN)
spellingShingle Chang-Yeh Hsieh
Ling-Shen Tseng
Yi-Han Chen
Chiung-Hui Tsai
Chih-Hung Wu
Deep-Learning Methods for Defect Inspection of Plated Through Holes With Clustering-Based Auto-Labeling and GAN-Based Model Training
IEEE Access
Automated optical inspection (AOI)
defect detection
deep learning
electroplated through-hole (PTH)
printed circuit board (PCB)
generative adversarial network (GAN)
title Deep-Learning Methods for Defect Inspection of Plated Through Holes With Clustering-Based Auto-Labeling and GAN-Based Model Training
title_full Deep-Learning Methods for Defect Inspection of Plated Through Holes With Clustering-Based Auto-Labeling and GAN-Based Model Training
title_fullStr Deep-Learning Methods for Defect Inspection of Plated Through Holes With Clustering-Based Auto-Labeling and GAN-Based Model Training
title_full_unstemmed Deep-Learning Methods for Defect Inspection of Plated Through Holes With Clustering-Based Auto-Labeling and GAN-Based Model Training
title_short Deep-Learning Methods for Defect Inspection of Plated Through Holes With Clustering-Based Auto-Labeling and GAN-Based Model Training
title_sort deep learning methods for defect inspection of plated through holes with clustering based auto labeling and gan based model training
topic Automated optical inspection (AOI)
defect detection
deep learning
electroplated through-hole (PTH)
printed circuit board (PCB)
generative adversarial network (GAN)
url https://ieeexplore.ieee.org/document/10792891/
work_keys_str_mv AT changyehhsieh deeplearningmethodsfordefectinspectionofplatedthroughholeswithclusteringbasedautolabelingandganbasedmodeltraining
AT lingshentseng deeplearningmethodsfordefectinspectionofplatedthroughholeswithclusteringbasedautolabelingandganbasedmodeltraining
AT yihanchen deeplearningmethodsfordefectinspectionofplatedthroughholeswithclusteringbasedautolabelingandganbasedmodeltraining
AT chiunghuitsai deeplearningmethodsfordefectinspectionofplatedthroughholeswithclusteringbasedautolabelingandganbasedmodeltraining
AT chihhungwu deeplearningmethodsfordefectinspectionofplatedthroughholeswithclusteringbasedautolabelingandganbasedmodeltraining