YOLOv4-MN3 for PCB Surface Defect Detection
Source abstract
Abstract
Surface defect detection for printed circuit board (PCB) is indispensable for managing PCB production quality. However, automatic detection of PCB surface defects is still a challenging task because, even within the same category of surface defect, defects present great differences in morphology and pattern. Although many computer vision-based detectors have been established to handle these problems, current detectors struggle to achieve high detection accuracy, fast detection speed and low memory consumption simultaneously. To address those issues, we propose a cost-effective deep learning (DL)-based detector based on the cutting-edge YOLOv4 to detect PCB surface defect quickly and efficiently. The YOLOv4 is improved upon with respect to its backbone network and the activation function in its neck/prediction network. The improved YOLOv4 is evaluated with a customized dataset, collected from a PCB factory. The experimental results show that the improved detector achieved a high performance, scoring 98.64% on mean average precision (mAP) at 56.98 frames per second (FPS), outperforming the other compared SOTA detectors. Furthermore, the improved YOLOv4 reduced the parameter space of YOLOv4 from 63.96 M to 39.59 M and the number of multiply-accumulate operations (Madds) from 59.75 G to 26.15 G.
Source: crossref · Quality: baseline
Research context
OpenAlex topics: Industrial Vision Systems and Defect Detection · Advanced Neural Network Applications · Infrastructure Maintenance and Monitoring
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
Affiliations
College of Engineering, South China Agricultural University, Guangzhou 510642, China
Funding
Natural Science Foundation of Guangdong, China
References (33)
Show source-provided references
1. Soldering defect detection in automatic optical inspection10.1016/j.aei.2019.101004 ↗
2. A review and analysis of automatic optical inspection and quality monitoring methods in electronics industry10.1109/ACCESS.2020.3029127 ↗
3. A machine vision based automatic optical inspection system for measuring drilling quality of printed circuit boards10.1109/ACCESS.2016.2631658 ↗
4. An efficient similarity measure approach for PCB surface defect detection10.1007/s10044-017-0640-9 ↗
5. Yuk, E.H., Park, S.H., Park, C.-S., and Baek, J.-G. (2018). Feature-Learning-Based Printed Circuit Board Inspection via Speeded-Up Robust Features and Random Forest. Appl. Sci., 8.10.3390/app8060932 ↗
6. Implementation of an automatic optical inspection system for solder quality classification of THT solder joints10.1109/TCPMT.2018.2864666 ↗
7. Machine vision based online detection of PCB defect10.1016/j.micpro.2020.103807 ↗
8. Zou, Z., Shi, Z., Guo, Y., and Ye, J. (2019). Object detection in 20 years: A survey. arXiv preprint, Available online: https://arxiv.org/abs/1905.05055.
9. A comprehensive overview of feature representation for biometric recognition10.1007/s11042-018-6808-5 ↗
10. Girshick, R., Donahue, J., Darrell, T., and Malik, J. (2014, January 23–28). Rich feature hierarchies for accurate object detection and semantic segmentation. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Columbus, OH, USA.10.1109/CVPR.2014.81 ↗
11. Girshick, R. (2015). Fast R-CNN. arXiv preprints, Available online: https://arxiv.org/abs/1504.08083.10.1109/ICCV.2015.169 ↗
12. Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks10.1109/TPAMI.2016.2577031 ↗
13. A comprehensive and systematic look up into deep learning based object detection techniques: A review10.1016/j.cosrev.2020.100301 ↗
14. Development of a YOLO-V3-based model for detecting defects on steel strip surface10.1016/j.measurement.2021.109454 ↗
15. Automatic defect detection of metro tunnel surfaces using a vision-based inspection system10.1016/j.aei.2020.101206 ↗
16. Deep learning-based fast recognition of commutator surface defects10.1016/j.measurement.2021.109324 ↗
17. TDD-net: A tiny defect detection network for printed circuit boards10.1049/trit.2019.0019 ↗
18. Detection of PCB Surface Defects with Improved Faster-RCNN and Feature Pyramid Network10.1109/ACCESS.2020.3001349 ↗
19. A Lightweight One-Stage Defect Detection Network for Small Object Based on Dual Attention Mechanism and PAFPN10.3389/fphy.2021.708097 ↗
20. Bochkovskiy, A., Wang, C.Y., and Liao, H. (2020). Yolov4: Optimal speed and accuracy of object detection. arXiv preprint, Available online: https://arxiv.org/abs/2004.10934.
21. Redmon, J., and Farhadi, A. (2018). Yolov3: An incremental improvement. arXiv preprint, Available online: https://arxiv.org/abs/1804.02767.
22. Spatial pyramid pooling in deep convolutional networks for visual recognition10.1109/TPAMI.2015.2389824 ↗
23. Liu, S., Qi, L., Qin, H., Shi, J., and Jia, J. (2018, January 18–23). Path Aggregation Network for Instance Segmentation. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Salt Lake City, UT, USA. Available online: https://ieeexplore.ieee.org/document/8579011.10.1109/CVPR.2018.00913 ↗
24. Wang, C.Y., Liao, H., Wu, Y.H., Chen, P.Y., and Yeh, I.H. (2020, January 14–19). CSPNet: A New Backbone that can Enhance Learning Capability of CNN. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, Seattle, WA, USA.10.1109/CVPRW50498.2020.00203 ↗
25. Howard, A., Sandler, M., and Chen, B. (November, January 27). Searching for mobileNetV3. Proceedings of the IEEE/CVF International Conference on Computer Vision, Seoul, Korea.
26. Misra, D. (2019). Mish: A self regularized non-monotonic neural activation function. arXiv preprint, Available online: https://arxiv.org/abs/1908.08681.
27. Real-time railroad track components inspection based on the improved yolov4 framework10.1016/j.autcon.2021.103596 ↗
28. Ramachandran, P., Zoph, B., and Le, Q.V. (2017). Searching for activation functions. arXiv preprint, Available online: https://arxiv.org/abs/1710.05941.
29. Deep sparse rectifier neural networks
30. Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classification
31. Eger, S., Youssef, P., and Gurevych, I. (November, January 31). Is it Time to Swish? Comparing Deep Learning Activation Functions across NLP tasks. Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing, Brussels, Belgium.10.18653/v1/D18-1472 ↗
32. Zhao, J., Zhang, X., Yan, J., Qiu, X., Yao, X., Tian, Y., Zhu, Y., and Cao, W. (2021). A Wheat Spike Detection Method in UAV Images Based on Improved YOLOv5. Remote Sens., 13.10.3390/rs13163095 ↗
33. Quantization friendly mobilenet (qf-mobilenet) architecture for vision based applications on embedded platforms10.1016/j.neunet.2020.12.022 ↗