A Performance Analysis of YOLOv8 and YOLOv10 on PCB Surface Defects

Abstract

The shift towards high-speed, automated optical inspection in electronics manufacturing necessitates object detection models that optimally balance accuracy with inference speed. This study presents a controlled, like-for-like performance evaluation of two successive models from the YOLO for the detection of six common surface defects on printed circuit boards (PCBs) using the PKU synthetic dataset. Under identical training protocols, hyperparameters, and data splits, both models achieved strong overall accuracy. YOLOv10 attained a mean Average Precision (mAP@50) of 0.893 and a stricter localization accuracy (mAP@50-95) of 0.460, slightly outperforming YOLOv8 attained 0.889 and 0.451, respectively. This marginal gain in accuracy, particularly at higher IoU thresholds, is contextualized within the established performance trade-off framework for real-time detectors, where architectural efficiency gains must be weighed against potential gains in precision. A class-wise analysis revealed that "missing_hole" defects were detected with the highest reliability, while classes like "mouse_bite" and "spur" presented challenges, primarily in recall. The results underscore YOLOv10's architectural improvements for efficient, high-precision detection. The discussion extends to the implications of these findings for industrial deployment, where the choice between model versions hinges on the specific production line's tolerance for false positives versus the imperative for millisecond-level latency. Future work should focus on recall improvement for minority defect classes and explicit measurement of the accuracy-speed Pareto frontier on target hardware to fully inform model selection for in-line AOI systems.

Keywords

• YOLOv8
• YOLOv10
• Surface Defect
• Printed Circuit Board
• Deep Learning

References

1. Budak, İ., Bal, S., Korkmaz, H., Üniversitesi, M., Fakültesi, T., Bölümü, E.-E. M., … Bölümü, T. (2025). PCB Üretiminde Çok Sınıflı Kusur Tespiti için YOLO Tabanlı Derin Öğrenme Modeli. Afyon Kocatepe Üniversitesi Fen Ve Mühendislik Bilimleri Dergisi, 25(4), 816–826. doi:10.35414/AKUFEMUBID.1551996
2. Chen, W., Huang, Z., Mu, Q., & Sun, Y. (2022). PCB Defect Detection Method Based on Transformer-YOLO. IEEE Access, 10, 129480–129489. doi:10.1109/ACCESS.2022.3228206 Coultard, F. (2005). Past, Present & Future. Circuit World, 31(2). doi:10.1108/CW.2005.21731BAF.001/FULL/XML
3. Ding, R., Dai, L., Li, G., & Liu, H. (2019). TDD-Net: A tiny defect detection network for printed circuit boards. CAAI Transactions on Intelligence Technology, 4(2), 110–116. doi:10.1049/TRIT.2019.0019
4. Elsharkawy, Z. F. (2025). Enhanced YOLOv11 framework for high precision defect detection in printed circuit boards. Scientific Reports 2025 15:1, 15(1), 42550-. doi:10.1038/s41598-025-27415-w
5. Ge, Z., Liu, S., Wang, F., Li, Z., & Sun, J. (2021). YOLOX: Exceeding YOLO Series in 2021, 5, 12. Retrieved from https://arxiv.org/pdf/2107.08430
6. GitHub - Ixiaohuihuihui/Tiny-Defect-Detection-for-PCB: This is a repository about PCB defect detection. (n.d.). Retrieved 13 August 2025, from https://github.com/Ixiaohuihuihui/Tiny-Defect-Detection-for-PCB
7. GitHub - tangsanli5201/DeepPCB: A PCB defect dataset. (n.d.). Retrieved 12 August 2025, from https://github.com/tangsanli5201/DeepPCB
8. Gündüz, M. Ş., & Işık, G. (2023a). A new YOLO-based method for real-time crowd detection from video and performance analysis of YOLO models. Journal of Real-Time Image Processing 2023 20:1, 20(1), 5-. doi:10.1007/S11554-023-01276-W

9. Gündüz, M. Ş., & Işık, G. (2023b). A new YOLO-based method for social distancing from real-time videos. Neural Computing and Applications 2023 35:21, 35(21), 15261–15271. doi:10.1007/S00521-023-08556-3
10. Huang, W., & Wei, P. (2019). A PCB Dataset for Defects Detection and Classification, 14(8). Retrieved from https://arxiv.org/pdf/1901.08204
11. Kaur, R., & Singh, S. (2023). A comprehensive review of object detection with deep learning. Digital Signal Processing, 132, 103812. doi:10.1016/J.DSP.2022.103812
12. Li, G., Gan, Y., Zhang, W., & Che, H. (2026). GS-YOLO: A lightweight and high-performance method for PCB surface defect detection. Expert Systems with Applications, 303, 130583. doi:10.1016/J.ESWA.2025.130583 Lim, J. Y., Lim, J. Y., Baskaran, V. M., & Wang, X. (2023). A deep context learning based PCB defect detection model with anomalous trend alarming system. Results in Engineering, 17, 100968. doi:10.1016/J.RINENG.2023.100968
13. Liu, Z., Lin, Y., Cao, Y., Hu, H., Wei, Y., Zhang, Z., … Guo, B. (2021). Swin Transformer: Hierarchical Vision Transformer Using Shifted Windows. In International Conference on Computer Vision (ICCV) (pp. 10012–10022). Retrieved from https://github.
14. Mazzetto, M., Teixeira, M., Oliveira Rodrigues, É., & Casanova, D. (2020). Deep learning models for visual inspection on automotive assembling line. Arxiv.OrgM Mazzetto, M Teixeira, ÉO Rodrigues, D CasanovaarXiv Preprint ArXiv:2007.01857, 2020•arxiv.Org, 7(4), 2456–1908. doi:10.22161/ijaers.74.56
15. Ren, Z., Fang, F., Yan, N., & Wu, Y. (2021). State of the Art in Defect Detection Based on Machine Vision. International Journal of Precision Engineering and Manufacturing-Green Technology 2021 9:2, 9(2), 661–691. doi:10.1007/S40684-021-00343-6
16. Shu, Y., Li, B., & Lin, H. (2021). Quality safety monitoring of LED chips using deep learning-based vision inspection methods. Measurement, 168, 108123. doi:10.1016/J.MEASUREMENT.2020.108123
17. Song, C., Zhou, Y., Ma, Y., Qi, Q., Wang, Z., & Hu, K. (2025). YOLO-HDEW: An Efficient PCB Defect Detection Model. Electronics 2025, Vol. 14, Page 3383, 14(17), 3383. doi:10.3390/ELECTRONICS14173383
18. Sundaresan Geetha, A., Alif, M. A. R., Hussain, M., & Allen, P. (2024). Comparative Analysis of YOLOv8 and YOLOv10 in Vehicle Detection: Performance Metrics and Model Efficacy. Vehicles 2024, Vol. 6, Pages 1364-1382, 6(3), 1364–1382. doi:10.3390/VEHICLES6030065
19. Tang, Y., Liu, R., & Wang, S. (2025). YOLO-SUMAS: Improved Printed Circuit Board Defect Detection and Identification Research Based on YOLOv8. Micromachines 2025, Vol. 16, Page 509, 16(5), 509. doi:10.3390/MI16050509
20. Tariq, M. F., & Javed, M. A. (2025). Small Object Detection with YOLO: A Performance Analysis Across Model Versions and Hardware. Arxiv. Retrieved from https://arxiv.org/pdf/2504.09900v1
21. Wang, A., Chen, H., Liu, L., Chen, K., Lin, Z., Han, J., & Ding, G. (2024). YOLOv10: Real-Time End-to-End Object Detection. In Advances in Neural Information Processing Systems (Vol. 37, pp. 107984–108011). Retrieved from https://github.com/THU-MIG/yolov10.
22. Wang, J., Huang, Z., Dong, • Yan, Hu, Y., & Zhao, H. (2025). LE-YOLO: a lightweight and effective YOLO model for remote sensing image object detection. Journal of Real-Time Image Processing 2025 23:1, 23(1), 27-. doi:10.1007/S11554-025-01822-8
23. Xiao, G., Hou, S., & Zhou, H. (2024). PCB defect detection algorithm based on CDI-YOLO. Scientific Reports, 14(1), 7351. doi:10.1038/S41598-024-57491-3
24. Yang, Y., & Kang, H. (2023). An Enhanced Detection Method of PCB Defect Based on Improved YOLOv7. Electronics 2023, Vol. 12, Page 2120, 12(9), 2120. doi:10.3390/ELECTRONICS12092120
25. Yi, F., Mohamed, A. S. A., Noor, M. H. M., Ani, F. C., & Zolkefli, Z. E. (2024). YOLOv8-DEE: a high-precision model for printed circuit board defect detection. PeerJ Computer Science, 10, e2548. doi:10.7717/PEERJ-CS.2548/SUPP-2
26. Zaidi, S. S. A., Ansari, M. S., Aslam, A., Kanwal, N., Asghar, M., & Lee, B. (2022). A survey of modern deep learning based object detection models. Digital Signal Processing, 126, 103514. doi:10.1016/J.DSP.2022.103514
27. Zhu, H., Huang, J., Liu, H., Zhou, Q., Zhu, J., & Li, B. (2022). Deep-Learning-Enabled Automatic Optical Inspection for Module-Level Defects in LCD. IEEE Internet of Things Journal, 9(2), 1122–1135. doi:10.1109/JIOT.2021.3079440