Synthesis of virtual reality and computer vision technologies in industrial radiography simulator

Purpose of research. Nowadays, the digitalization of production is considered as the most important aspect of technological growth to improve the competitiveness of enterprises. An innovative approach combining virtual reality and computer vision technologies into a single tool designed to improve the quality of practice-oriented training in the field of industrial radiography is proposed. Within the framework of the article the research of the most effective models of artificial neural networks in application to the task of detection of defective areas of welded metal joints on radiographic images is carried out. A detailed analysis of the YOLOv8 architecture with respect to the detection of small-sized defects is carried out. A method for synthesizing virtual reality and computer vision technologies in a single educational tool for industrial radiography is described.

Methods. Methods of empirical research, system analysis and synthesis of related information technologies were used in this work.

Results. The empirical study revealed the limited effectiveness of the YOLOv10 model as applied to the generalization of features of objects of small dimensionality and low contrast. YOLOv8 showed more practical results and greater stability when generalizing the contour component of defects. In the process of system analysis of YOLOv8 architecture the loss of spatial information when using sequential convolutional operations preceding upsampling was revealed. Modification of the basic YOLOv8 architecture was performed in order to improve the generalization ability of lowdimensional and low-contrast defects. The methodology of synthesis of virtual reality and computer vision technologies in the form of an intelligent assistant for intellectualization of nondestructive testing process is presented.

Conclusion. The integration of the above synthesis method into a single software product will improve the quality of specialist training and open access to innovative methods of improving professional skills at every stage of a professional career.

1. Blagodeteleva N.K. Trends and prospects of using modern VR technologies in foreign language education in the Russian higher education system. Mir nauki, kul'tury, obrazovaniya = The World of Science, Culture, and Education. 2023; (3):109–113. (In Russ.)

2. Mirgorodskaya O.N. A study of the experience of using immersive technologies in the marketing activities of retail enterprises. In: Predprinimatel'stvo, marketing i logistika v tsifrovoi ekonomike: materialy II Vserossiiskoi konferentsii = Entrepreneurship, marketing and logistics in the digital economy: Proceedings of the II All-Russian Conference. Orel: Orlovskii gosudarstvennyi universitet imeni I.S. Turgeneva; 2024. P. 183–190. (In Russ.)

3. Rysbai N. Machine learning in advanced research. development processes. Vestnik nauki = Bulletin of Science. 2023;3(5):604–611. (In Russ.)

4. Ozdamirov R.V. The influence of virtual reality technologies on learning in a digital educational environment. In: Badelina M.V. (ed.) Perspektivy razvitiya vysshei shkoly: materialy I Mezhdunarodnoi nauchno-prakticheskoi konferentsii = Prospects for the development of higher education: Materials of the I International Scientific and Practical Conference. Tyumen: Tyumen Industrial University; 2020. P. 250–255. (In Russ.)

5. Garaev A.I. Development of educational VR technologies in the digital environment. In: Maltsev N.V. (ed.) Konkurentosposobnost' sub"ektov khozyaistvovaniya v usloviyakh novykh vyzovov vneshnei sredy: problemy i puti ikh resheniya: Sbornik materialov Mezhdunarodnoi nauchno-prakticheskoi konferentsii = Competitiveness of business entities in the context of new challenges of the external environment: problems and ways to solve them: A collection of materials of the International scientific and practical conference. Yekaterinburg: Izdatel'stvo Ural'skogo gosudarstvennogo gornogo universiteta; 2020. P. 20. (In Russ.)

6. Surai N.M., Teplaya N.A. Modern trends in the application of digital technologies in Russian retail. Ekonomika. Professiya. Biznes = Economy. Profession. Business. 2024;(2):94– 104. (In Russ.)

7. Gusakova A.A., Kalvinkovskaya V.D., Baibardina T.N. Prospects of using machine learning and artificial intelligence technologies in marketing. (In Russ.) Available at: http://lib.i-bteu.by/bitstream/handle/22092014/6382/Гусакова%20А.А._Кальвинковская%20B.D.%20Perspectives.pdf?sequence=1&isAllowed=y (accessed 09.11.2024).

8. Polevoda I.I., et al. Virtual and augmented reality technologies in the educational process. Vestnik Universiteta grazhdanskoi zashchity MChS Belarusi = Bulletin of the University of Civil Protection of the Ministry of Emergency Situations of Belarus. 2022;6(1);119–142. (In Russ.)

9. Trampus P. Symbiosis of industry 4.0 and non-destructive evaluation. In: Buletin Scientific. Kishinev; 2023. P. 43‒51.

10. Saboonchi H., Blanchette D., Hayes K. Advancements in radiographic evaluation through the migration into NDE 4.0. Journal of nondestructive evaluation = Journal of Nondestructive Evaluation. 2021;40:1–12.

11. Meyendorf N.G., Ida N., Singh R., Vrana J. NDE 4.0: Progress, promise, and its role to industry 4.0. NDT & E International. 2023;(140):102957. https://doi.org/10.1016/j.ndteint.2023.102957

12. Nizhnik A.E., et al. Classification of modern methods of non-destructive testing. The possibility of their application for diagnostics of equipment in the oil and gas industry. Stroitel'stvo neftyanykh i gazovykh skvazhin na sushe i na more = Construction of Oil and Gas Wells on Land and at Sea. 2020;(5):41–46. (In Russ.)

13. Kuvshinnikov V.S., Kovshov E.E. Training specialists in radiation non-destructive testing using digital technologies. Izvestiya Yugo-Zapadnogo gosudarstvennogo universiteta. Seriya: Upravlenie, vychislitel'naya tekhnika, informatika. Meditsinskoe priborostroenie = Proceedings of the Southwest State University. Series: Control, Computer Engineering, Information Science. Medical Instruments Engineering. 2024;13(4):54–69. (In Russ.)

14. Sohan M., et al. A review on YOLOv8 and its advancements. In: International Conference on Data Intelligence and Cognitive Informatics. Singapore: Springer; 2024. P. 529– 545.

15. Korchagin V.D. Analysis of modern SOTA architectures of artificial neural networks for solving problems of image classification and object detection. Programmnye sistemy i vychislitel'nye metody = Software Systems and Computational Methods. 2023;(4):73–87. (In Russ.)

16. Wang A., et al. Yolov10: Real-time end-to-end object detection. Available at: https://arxiv.org/abs/2405.14458 (accessed 09.11.2024).

17. Korchagin V.D., Kuvshinnikov V.S., Kovshov E.E. Criterion analysis of data processing models for radiation non-destructive testing. International Journal of Open Information Technologies. 2024;12(4):23–31.

18. Qilong Wang, Banggu Wu, Pengfei Zhu, Peihua Li, Wangmeng Zuo, Qinghua Hu. ECA-Net: Efficient Channel Attention for Deep Convolutional Neural Networks. Available at: https://arxiv.org/abs/1910.03151 (accessed 11.09.2024).

19. Wang Н., Song X. DC-YDLO: A dual channel YOLO for small object defect detection of circuit boards. In: 2024 5th International Conference on Computer i Vision, Image and Deep Learning (CVIDL). IEEE; 2024. P. 1292–1296.

20. Liang-Chieh Chen, Papandreou G., Kokkinos I., Murphy K., Yuille A. L. DeepLab: Semantic Image Segmentation with Deep Convolutional Nets, Atrous Convolution, and Fully Connected CRFs. IEEE Trans. Pattern. Anal. Mach. Intell. 2018;(40):834‒848. https://doi.org/10.1109/TPAMI.2017.2699184

21. Larobina M. Thirty years of the DICOM standard. Tomography. 2023;9(5):1829– 1838.

22. Vrana J. Industrial Internet of things, digital twins, and cyber-physical loops for NDE 4.0. In: Handbook of Nondestructive Evaluation 4.0. Cham: Springer; 2022. P. 295–328.