A Comparative Analysis of Printed Circuit Boards Defect Detection Leveraging Deep Learning Approaches

Zihan Zhou

doi:10.54097/11xrk596

Authors

Zihan Zhou

DOI:

https://doi.org/10.54097/11xrk596

Keywords:

Deep learning; printed circuit boards; defect detection.

Abstract

Printed circuit boards (PCBs) are the foundations of modern industry. Their prevalence across electronic industries is explicit, ranging from computers and smartphones to televisions, digital cameras, and other devices. However, during the production process of PCBs, defects are inevitable, leading to substandard products, hence the detection of defects is a crucial task. Deep learning offers significant advantages over conventional machine learning approaches. These advantages include but not limited to the capacity to automatically extract complex embeddings from input samples, leading to improved accuracy and performance. Fueled deep learning, algorithms designed for PCB defect detection are anticipated to deliver superior precision and efficiency, attributed to their capability to recognize novel defect patterns. This article provides a comprehensive analysis of these algorithms, separated into three major categories, including Transformer-based solutions, one-stage, and two-stage approaches. Improvement methods and advantages of these algorithms are detailed. Finally, prospective future research is pinpointed to tackle existing challenges within the algorithms.

Downloads

Download data is not yet available.

References

[1] Jiang Wujin, Shaolin Zhang, Wenbin Chen. Multi-target Detection of PCB Defects Based on Improved SSD. International Core Journal of Engineering, 2022, 8(6): 319-325.

[2] Zhang Yue, Xie Fei, Huang Lei, et al. A lightweight one-stage defect detection network for small object based on dual attention mechanism and PAFPN. Frontiers in Physics, 2021, 9: 708097. DOI: https://doi.org/10.3389/fphy.2021.708097

[3] Lin Tsung-Yi, Goyal Priya, Girshick Ross, et al. Focal loss for dense object detection. Proceedings of the IEEE international conference on computer vision. 2017: 2980-2988. DOI: https://doi.org/10.1109/ICCV.2017.324

[4] Li Jing, Li Weiye, Chen Yingqian, et al. Research on object detection of PCB assembly scene based on effective receptive field anchor allocation. Computational Intelligence and Neuroscience, 2022, 2022(1): 7536711. DOI: https://doi.org/10.1155/2022/7536711

[5] Luo Wenjie, Li Yujia, Urtasun Raquel, et al. Understanding the effective receptive field in deep convolutional neural networks. Advances in neural information processing systems, 2016, 29: 1-9.

[6] Chen Liang-Chieh, Papandreou George, Kokkinos Iasonas, et al. Deeplab: Semantic image segmentation with deep convolutional nets, atrous convolution, and fully connected crfs. IEEE transactions on pattern analysis and machine intelligence, 2017, 40(4): 834-848. DOI: https://doi.org/10.1109/TPAMI.2017.2699184

[7] Chen Liang-Chieh, Papandreou George, Schroff Florian, et al. Rethinking atrous convolution for semantic image segmentation. arXiv preprint, 2017: 1706.05587.

[8] Wu Ligang, Liang Zhang, Qian Zhou. Printed circuit board quality detection method integrating lightweight network and dual attention mechanism. IEEE Access, 2022, 10: 87617-87629. DOI: https://doi.org/10.1109/ACCESS.2022.3198994

[9] Xuan Wang, Jian-She GAO, Bo-Jie Hou, et al. A lightweight modified YOLOX network using coordinate attention mechanism for PCB surface defect detection. IEEE sensors journal, 2022, 22(21): 20910-20920. DOI: https://doi.org/10.1109/JSEN.2022.3208580

[10] Zhao YiHang, HuiChen Yang, HaiHong Feng. An improved YOLOv5 PCB defect detection. International Conference on Advanced Sensing and Smart Manufacturing, 2022, 12351: 380-387. DOI: https://doi.org/10.1117/12.2652341

[11] Liu Songtao, Di Huang, Yunhong Wang. Learning spatial fusion for single-shot object detection. arXiv preprint, 2019: 1911.09516.

[12] Liu Yichao, Zongru Shao, Nico Hoffmann. Global attention mechanism: Retain information to enhance channel-spatial interactions. arXiv preprint, 2021: 2112.05561.

[13] Zheng Jianfeng, Sun Xiaopeng, Zhou Haixiang, et al. Printed circuit boards defect detection method based on improved fully convolutional networks. IEEE Access, 2022, 10: 109908-109918. DOI: https://doi.org/10.1109/ACCESS.2022.3214306

[14] Yu Zhilong, Wu Yuxiang, Wei Binqian, et al. A lightweight and efficient model for surface tiny defect detection. Applied Intelligence, 2023, 53(6): 6344-6353. DOI: https://doi.org/10.1007/s10489-022-03633-x

[15] Ding Runwei, Dai Linhui, Li Guangpeng, et al. TDD‐net: a tiny defect detection network for printed circuit boards. CAAI Transactions on Intelligence Technology, 2019, 4(2): 110-116. DOI: https://doi.org/10.1049/trit.2019.0019

[16] Shrivastava Abhinav, Abhinav Gupta, Ross Girshick. Training region-based object detectors with online hard example mining. Proceedings of the IEEE conference on computer vision and pattern recognition. 2016: 761-769. DOI: https://doi.org/10.1109/CVPR.2016.89

[17] Hu Bing, Jianhui Wang. Detection of PCB surface defects with improved faster-RCNN and feature pyramid network. Ieee Access, 2020, 8: 108335-108345. DOI: https://doi.org/10.1109/ACCESS.2020.3001349

[18] Li Dan, Fu SongLing, Zhang QiJun, et al. An improved PCB defect detector based on feature pyramid networks. International Conference on Computer Science and Artificial Intelligence. 2020: 232-239. DOI: https://doi.org/10.1145/3445815.3445853

[19] Touvron Hugo, Cord Matthieu, Douze Matthijs, et al. Training data-efficient image transformers & distillation through attention. International conference on machine learning. PMLR, 2021: 10347-10357.

[20] Dosovitskiy Alexey, Beyer Lucas, Kolesnikov Alexander, et al. An image is worth 16x16 words: Transformers for image recognition at scale. arXiv preprint, 2020: 2010.11929.

[21] A Kang, Yanping Zhang. LPViT: a transformer based model for PCB image classification and defect detection. IEEE Access, 2022, 10: 42542-42553. DOI: https://doi.org/10.1109/ACCESS.2022.3168861

[22] Chen Wei, Huang Zhongtian, Mu Qian, et al. PCB defect detection method based on transformer-YOLO. IEEE Access, 2022, 10: 129480-129489. DOI: https://doi.org/10.1109/ACCESS.2022.3228206