Deep Learning Application for Image-Based Defect Detection in 3D Printing Processes

Abstract

Fused Deposition Modeling (FDM) is a widely preferred additive manufacturing technology due to its design flexibility and cost advantages. However, structural and visual defects such as stringing, layer shifting, warping, or off-platform issues frequently occur in FDM-based three-dimensional (3D) printing processes. This results in losses of raw materials, energy, and time. This study aims to classify common defect types occurring in FDM-based 3D printing using image-based deep learning algorithms. In this study, a transfer learning approach is adopted using the EfficientNetB0 and MobileNetV2 architectures, both pre-trained on ImageNet, with a dataset of 1912 images. The model’s generalization capability is enhanced through data preprocessing and enhancement techniques. The model can successfully classify defect types using print images without additional sensors or hardware components. The results show that the EfficientNetB0 model has an overall accuracy of 87.7%, while MobileNetV2 achieves 97%. The MobileNetV2 architecture demonstrates strong performance, particularly in error classes such as Layer Shifting and Stringing, with high F1 scores. The proposed architectures have the potential to reduce losses in FDM-based 3D printing processes by providing a low-cost and accessible visual quality control system.

Keywords

• 3D printing
• deep learning
• defect detection
• image processing

Thanks

The authors would like to acknowledge the use of the FDM 3D Printing Defect Dataset obtained from Kaggle for the experimental phase of this research. This work is derived from the doctoral thesis entitled ‘Development of a Hybrid Machine Learning Method for Quality Optimization in 3D Printing’. The study was supported by the Afyon Kocatepe University Scientific Research Projects Coordination Unit (BAP) under Project No: 25.FEN.BİL.11.

3B Baskı Süreçlerinde Görüntü Tabanlı Hata Tespiti için Derin Öğrenme Uygulaması

Abstract

Erimiş Katmanlı Modelleme (FDM), tasarım esnekliği ve maliyet avantajları nedeniyle yaygın olarak tercih edilen bir eklemeli üretim teknolojisidir. Bununla birlikte, FDM tabanlı üç boyutlu (3B) baskı süreçlerinde sıklıkla ipliklenme, katman kayması, bükülme veya platform dışı sorunlar gibi yapısal ve görsel kusurlar meydana gelir. Bu durum, hammadde, enerji ve zaman kayıplarına yol açar. Bu çalışma, görüntü tabanlı derin öğrenme algoritmaları kullanarak FDM tabanlı 3B baskıda meydana gelen yaygın kusur türlerini sınıflandırmayı amaçlamaktadır. Bu çalışmada, 1912 görüntüden oluşan bir veri kümesiyle ImageNet üzerinde önceden eğitilmiş EfficientNetB0 ve MobileNetV2 mimarileri kullanılarak bir transfer öğrenme yaklaşımı benimsenmiştir. Modelin genelleme yeteneği, veri ön işleme ve geliştirme teknikleri ile artırılmıştır. Model, ek sensörlere veya donanım bileşenlerine ihtiyaç duymadan, yalnızca baskı görüntülerini kullanarak kusur türlerini başarıyla sınıflandırabilir. Sonuçlar, EfficientNetB0 modelinin %87,7'lik genel doğruluk oranına sahip olduğunu, MobileNetV2'nin ise %97'lik bir doğruluk oranına ulaştığını göstermektedir. MobileNetV2 mimarisi, özellikle Katman Kayması ve İplikleme gibi hata sınıflarında yüksek F1 puanlarıyla güçlü bir performans sergilemektedir. Bu çalışmada önerilen başarılı mimariler, gelecekteki çalışmalarda ek donanım bileşenlerine ihtiyaç duymadan düşük maliyetli ve erişilebilir bir görsel kalite kontrol sistemi sağlayarak FDM tabanlı 3D baskı süreçlerindeki kayıpları azaltma potansiyeline sahiptir.

Keywords

• 3B Baskı
• derin öğrenme
• hata tespiti
• görüntü işleme

Thanks

Yazarlar, bu araştırmanın deneysel aşamasında Kaggle'dan elde edilen FDM 3D Baskı Hata Veri Seti'nin kullanımını kabul etmektedir. Bu çalışma, "3D Baskıda Kalite Optimizasyonu için Hibrit Makine Öğrenme Yönteminin Geliştirilmesi" başlıklı doktora tezinden türetilmiştir. Çalışma, Afyon Kocatepe Üniversitesi Bilimsel Araştırma Projeleri Koordinasyon Birimi (BAP) tarafından 25.FEN.BİL.11 numaralı proje kapsamında desteklenmiştir.

References

1. Abdulshahed, A. M., & Wafa, F. (2025). Surface roughness prediction in additive manufacturing: presenting the power of neural networks compared to linear regression. Journal of Advanced Manufacturing Systems, 24(01), 69-88.
2. Aktepe, E., & Aktepe, Ş. (2024). PLA ve geri dönüştürülmüş PET Filamentlerinin 3D Fdm Baskida Boyutsal Doğruluk ve Geriçekilme Performansinin Karşilaştirilmasi. International Journal of 3D Printing Technologies and Digital Industry, 8(1), 114-123.
3. Aktepe, E., & Ergün, U. (2025a). A Systematic Analysis of 3D Printing Research in Doctoral and Specialization Theses in Turkey. In S. Kocer & O. Dundar (Eds.), Next generation engineering: Smart solutions and applications (pp. 13–36). ISRES Publishing
4. Aktepe, E., & Ergün, U. (2025b). Machine learning approaches for FDM-based 3D printing: a literature review. Applied Sciences, 15(18), 10001.
5. Aktepe, E., & Koca, Y. B. (2024). Optimization of 3D Printing Parameters Using Machine Learning Techniques. In International Congress of Electrical and Computer Engineering (pp. 277-286). Cham: Springer Nature Switzerland.
6. Ali, S., Nouzil, I., Mehra, V., Eltaggaz, A., Deiab, I., & Pervaiz, S. (2025). Integrated optimization scheme for 3D printing of PLA-APHA biodegradable blends. Progress in Additive Manufacturing, 10(1), 875-886.
7. Alli, Y. A., Anuar, H., Manshor, M. R., Okafor, C. E., Kamarulzaman, A. F., Akçakale, N., Mohd Nazeri, F. N., Bodaghi, M., Suhr, J., & Nasir, N. A. M. (2024). Optimization of 4D/3D printing via machine learning: A systematic review. Hybrid Advances, 6, 100242.
8. Armin, E., Ebrahimian, S., Sanjari, M., Saidi, P., & Pourreza, H. R. (2025). Defect detection in 3D printing: A review of image processing and machine vision techniques. The International Journal of Advanced Manufacturing Technology, 140(3), 2103-2128.

9. Badour, S., & Novák, M. (2024). Real-time vision-based fault detection system for FDM 3D printing using convolutional neural networks. In 2024 21st International Conference on Mechatronics-Mechatronika (ME) (pp. 1-9). IEEE.
10. Baek, S. (2024). Multivariate sensor data analysis for fault detection toward feedback loop-based continuous learning in FFF 3D printer. The International Journal of Advanced Manufacturing Technology, 132(11), 5975-5988.
11. Bhandarkar, V. V., Karnati, M., & Tandon, P. (2025a). Defect detection in 3D-printed polymer parts using deep learning models: a comparative investigation. Rapid Prototyping Journal, 31(7), 1428-1448.
12. Bhandarkar, V. V., Kumar, A., & Tandon, P. (2025b). Warpage detection in 3D printing of polymer parts: a deep learning approach. Journal of Intelligent Manufacturing, 36(5), 3129-3141.
13. Bhandarkar, V. V., Shahare, H. Y., Mall, A. P., & Tandon, P. (2025c). An overview of traditional and advanced methods to detect part defects in additive manufacturing processes. Journal of Intelligent Manufacturing, 36(7), 4411-4446.
14. Błaszczykowski, M., Majerek, D., Sędzielewska, E., Tomiło, P., & Pytka, J. (2025). A novel machine learning system for early defect detection in 3D printing. Advances in Science and Technology. Research Journal, 19(3).
15. Çelebi, A., & İmanç, M. M. (2023). Evaluation of mechanical properties of PLA auxetic structures produced by additive manufacturing. Journal of Materials and Mechatronics: A, 4(2), 384-396.
16. Da Silva, L. M., Alcalá, S. G., de A. Barbosa, T. M. G., & Araújo, R. (2024). Object and defect detection in additive manufacturing using deep learning algorithms. Production Engineering, 18(6), 889-902.
17. Khan, M. F., Alam, A., Siddiqui, M. A., Alam, M. S., Rafat, Y., Salik, N., & Al-Saidan, I. (2021). Real-time defect detection in 3D printing using machine learning. Materials Today: Proceedings, 42, 521-528.
18. FDM 3D Printing Defect Dataset (2024). https://www.kaggle.com/datasets/wengmhu/fdm-3d-printing-defect-dataset
19. Goh, G. D., Sing, S. L., & Yeong, W. Y. (2021). A review on machine learning in 3D printing: applications, potential, and challenges. Artificial Intelligence Review, 54(1), 63–94.
20. Hu, W., Chen, C., Su, S., Zhang, J., & Zhu, A. (2024). Real-time defect detection for FFF 3D printing using lightweight model deployment. The International Journal of Advanced Manufacturing Technology, 134(9), 4871-4885.
21. Kadam, V., Kumar, S., Bongale, A., Wazarkar, S., Kamat, P., & Patil, S. (2021). Enhancing surface fault detection using machine learning for 3D printed products. Applied System Innovation, 4(2), 34.
22. Kadhum, A. H., Al-Zubaidi, S., & AlKareem, S. S. A. (2023). Optimization of mechanical properties and surface characteristics of PLA+ 3D printing materials. International Journal of Chemical Engineering, 2023(1), 8887905.
23. Kuriachen, B., Jeyaraj, R., Raphael, D., Ashok, P., Sundari, P. S., & Paul, A. (2025). Defect detection in fused deposition modelling using lightweight convolutional neural networks. Engineering Applications of Artificial Intelligence, 141, 109802.
24. Ngo, T. D., Kashani, A., Imbalzano, G., Nguyen, K. T., & Hui, D. (2018). Additive manufacturing (3D printing): A review of materials, methods, applications and challenges. Composites Part B: Engineering, 143, 172-196.
25. Paraskevoudis, K., Karayannis, P., & Koumoulos, E. P. (2020). Real-time 3D printing remote defect detection (stringing) with computer vision and artificial intelligence. Processes, 8(11), 1464.
26. Rajan, K., Samykano, M., Kadirgama, K., Harun, W. S. W., & Rahman, M. M. (2022). Fused deposition modeling: process, materials, parameters, properties, and applications. The International Journal of Advanced Manufacturing Technology, 120(3), 1531-1570.
27. Rettenberger, L., Beyer, N., Sieber, I., & Reischl, M. (2024). Fault detection in 3D-printing with deep learning. In 2024 IEEE International Conference on Consumer Electronics (ICCE) (pp. 1-4). IEEE.
28. Rojek, I., Mikołajewski, D., Kempiński, M., Galas, K., & Piszcz, A. (2025). Emerging applications of machine learning in 3D printing. Applied Sciences, 15(4), 1781.
29. Soori, M., Jough, F. K. G., Dastres, R., & Arezoo, B. (2025). Additive manufacturing modification by artificial intelligence, machine learning, and deep learning: A review. Additive Manufacturing Frontiers, 4(2), 200198.
30. Su, X., Peng, X., Zhou, X., Cao, H., Shan, C., Li, S., Qiao, S., & Shi, F. (2025, June). Enhanced defect detection in additive manufacturing via virtual polarization filtering and deep learning optimization. Photonics 12(6), 599.
31. Tamir, T. S., Xiong, G., Fang, Q., Yang, Y., Shen, Z., Zhou, M., & Jiang, J. (2023). Machine-learning-based monitoring and optimization of processing parameters in 3D printing. International Journal of Computer Integrated Manufacturing, 36(9), 1362-1378.
32. Ukwaththa, J., Herath, S., & Meddage, D. P. P. (2024). A review of machine learning (ML) and explainable artificial intelligence (XAI) methods in additive manufacturing (3D Printing). Materials Today Communications, 41, 110294.
33. Verana, M., Nwakanma, C. I., Lee, J. M., & Kim, D. S. (2021). Deep learning-based 3d printer fault detection. In 2021 Twelfth International Conference on Ubiquitous and Future Networks (ICUFN) (pp. 99-102). IEEE.
34. Wen, H., Huang, C., & Guo, S. (2021). The application of convolutional neural networks (CNNs) to recognize defects in 3D-printed parts. Materials, 14(10), 2575.
35. Wickramasinghe, S., Do, T., & Tran, P. (2020). FDM-based 3D printing of polymer and associated composite: A review on mechanical properties, defects and treatments. Polymers, 12(7), 1529.
36. Yan, J. (2020). 3D printing optimization algorithm based on back-propagation neural network. Journal of Engineering, Design and Technology, 18(5), 1223-1230.
37. Yeh, C. R., Yan, S. Q., & Tsai, M. J. (2023). YOLOv5 based Intelligent Defect Detection of Dual-color Gradient Object Fabricated by Fused Deposition Modeling Additive Manufacturing System. In 2023 International Automatic Control Conference (CACS) (pp. 1-6). IEEE.
38. Zhang, X., Chu, D., Zhao, X., Gao, C., Lu, L., He, Y., & Bai, W. (2024). Machine learning-driven 3D printing: A review. Applied Materials Today, 39, 102306.
39. Zhang, Y., Lin, C., Tian, Y., Gao, J., Song, B., Zhang, H., Wang, M., Song, K., Deng, B., Xue, D., Yao, Y., Shi, Y., & Fu, K. K. (2025). Machine learning enhanced metal 3D printing: high throughput optimization and material transfer extensibility. International Journal of Extreme Manufacturing, 7(4), 045004.