Model of an automated product quality control system based on neural networks with unsupervised and semi-supervised learning

This paper addresses the problem of automated visual inspection of welded and hull structures in shipbuilding, where product reliability and operational safety critically depend on the timely and accurate detection of defects. Traditional optical inspection is limited by subjective human assessment and poor scalability, while conventional computer vision techniques suffer from data scarcity and variability of industrial environments. To overcome these limitations, an integrated mathematical framework is proposed that combines unsupervised and semi-supervised learning approaches. The architecture includes: (i) a convolutional autoencoder trained on defect-free reference samples for reconstruction-based anomaly detection; (ii) a two-stage sliding-window algorithm with dual thresholds for distinguishing weak defects from background noise under controlled false alarm rates; and (iii) a semi-supervised classification module that integrates contrastive learning with graph-based pseudo-labeling methods (k-NN and label propagation) to leverage large-scale unlabeled datasets. Joint optimization of reconstruction and discriminative representation aligns normality criteria with stable classification boundaries. Experimental validation confirms that the proposed method reliably detects both prominent and subtle defects, minimizes dependence on manual labeling, and can be seamlessly integrated into industrial quality assurance workflows. The main contribution lies in the development of a unified inspection model that fuses reconstruction-based, contrastive, and graph-driven approaches, demonstrating potential for improving reproducibility, reducing labor intensity, and enhancing the reliability of shipbuilding production.

1. Milushkov, V. I., N. N. Limanskiy and V. E. Marley. “Integrating hybrid semi-supervised and contrastive learning for automatic defect classification in manufacturing data: lmproving product quality control accuracy.” Science Prospects 6(177) (2024): 81–86.

2. Stebakov, I. N., A. V. Kornaev and E. P. Kornaeva. “Intelligent diagnostics of rotary systems failures.” Energo — i resursosberezhenie — XXI vek: Materialy XIX mezhdunarodnoj nauchno-prakticheskoj konferencii, Oryol, 10–12 noyabrya 2021 goda / Pod redakciej A. N. Kachanova, YU.S. Stepanova. Orel: Orlovskiy gosudarstvennyy universitet imeni I. S. Turgeneva, 2021: 100–105.

3. Schetinin, E. Yu. “On automatic detection of anomalies in electrocardiogramms with generative machine learning.” Vestnik Komp’Iuternykh I informatsionnykh Tekhnologii 19.6(216) (2022): 51–59. DOI: 10.14489/vkit.2022.06.pp.051-059.

4. Biryulin, K. G. “Neural network training methods.” Sovremennoe professional’noe obrazovanie: opyt, problemy, perspektivy: Materialy VIII Mezhdunarodnoy nauchno-prakticheskoy konferentsii. V 2-kh chastyakh, Rostov-na-Donu, 22 marta 2021 goda. Tom Chast’ 1. Rostov-na-Donu: Yuzhnyy universitet (IUBiP) “Izdatel’stvo VVM”, 2021: 53–56.

5. Mukhachev, P. A., T. R. Sadretdinov et al. “Modern machine learning methods for telemetry-based spacecraft health monitoring.” Avtomatika I Telemehanika 8 (2021): 3–38. DOI: 10.31857/S0005231021080018.

6. Milushkov, V. I. “Application of automatic anomaly detection methods in automated industrial image inspection: opportunities and perspectives.” Aktual’nye resheniya problem vodnogo transporta: sbornik materialov III Mezhdunarodnoy nauchno-prakticheskoy konferentsii., Astrakhan’, 29–31 maya 2024 goda. Astrakhan’: Volzhskiy gosudarstvennyy universitet vodnogo transporta, 2024: 85–88.

7. Milushkov, V. I. “Obzor primeneniya metodov poluavtomaticheskogo i neupravlyaemogo obucheniya dlya povysheniya effektivnosti avtomaticheskogo promyshlennogo kontrolya izobrazheniy.” 74-ya Mezhdunarodnaya studencheskaya nauchno-tekhnicheskaya konferentsiya: Materialy, Astrakhan’, 08–13 aprelya 2024 goda. Astrakhan’: Astrakhanskiy gosudarstvennyy tekhnicheskiy universitet, 2024: 597–598.

8. Tsyba, E. N., O. A. Volkova and N. A. Vostrukhov. “Experience of using autoencoders in solving problems of detecting anomalies in time series of measurement information.” Al”Manac of Modern Metrology 2(38) (2024): 150–160.

9. He, K., R. Girshick, et al. “Momentum Contrast for Unsupervised Visual Representation Learning.” Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)2020.

10. He, K., X. Zhang, S. Ren and J. Sun. “Deep Residual Learning for Image Recognition.” Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR)2016.

11. Hinton, G. E., S. Osindero and Y-W. Teh. “A Fast Learning Algorithm for Deep Belief Nets.” Neural Computation 18.7 (2006): 1527–1554. DOI: 10.1162/neco.2006.18.7.1527.

12. Hu, G., X. Huang, et al. “Unsupervised fabric defect detection based on a deep convolutional generative adversarial network.” Textile Research Journal 90.3–4 (2020): 247–270. DOI: 10.1177/0040517519862880.

13. Tan, Y., A. Sangiovanni-Vincentelli, et al. “An Encoder-Decoder Based Approach for Anomaly Detection with Application in Additive Manufacturing.” 2019 18th IEEE International Conference On Machine Learning And Applications (ICMLA) — 2019: 1008–1015. DOI: 10.1109/ICMLA.2019.00171.

14. Oh, S., M. Jung, C. Lim and S. Shin. “Automatic Detection of Welding Defects Using Faster R-CNN.” Applied Sciences 10.23 (2020). DOI: 10.3390/app10238629.

15. Wan, Y., L. Gao, X. Li and Y. Gao. “Semi-Supervised Defect Detection Method with Data-Expanding Strategy for PCB Quality Inspection.” Sensors 22.20 (2022). DOI: 10.3390/s22207971.

16. Wang, X., F. He and X. Huang. “A new method for deep learning detection of defects in X-ray images of pressure vessel welds.” Scientific Reports 14.1 (2024): 6312.