Russian Federation
Russian Federation
Abstract: The paper presents the results of a study into the application of YOLOv11 convolutional neural network model configurations to object segmentation tasks. It is hypothesized that YOLOv11 will be utilized in the domain of machine vision within the context of an automated railway rolling stock control system. The primary focus of this study was to assess the performance and quality of image processing implemented by various model configurations. A comparative analysis of 25 YOLOv11 configurations was conducted, varying in architecture and input image resolution within the range of 640 × 640 and 1920 × 1920 pixels. The creation of a specialized visual dataset was instrumental in facilitating the training of the models.This dataset comprised 20,000 annotated images of railway infrastructure, systematically distributed across 40 object classes. The performance and segmentation accuracy of all trained models was evaluated using the mAP metric (0.5–0.95). The results obtained are informative when choosing the configuration of theYOLOv11 model with the most suitable parameters for use in on-board machine vision systems of automated control systems, depending on the requirements for the accuracy of object segmentation and the power of available computing resources.
machine vision; automated control system; railway rolling stock; object detection; object segmentation; convolutional neural network; YOLOv11
1. Ohotnikov A. L. Proekty razvitiya sistem avtomatiche- skogo upravleniya dvizheniem poezdov / A. L. Ohotnikov, I. A. Volkova // Intellektual'nye sistemy i tehnolo- gii na transporte. — 2023. — № 1(25). — S. 25–31.
2. Ivanov V. F. Gradusov A. N. Algoritm kompleksirovaniya sensornyh dannyh dlya zadach avtomaticheskogo upravle- niya podvizhnym sostavom / V. F. Ivanov, A. L. Ohotnikov, A. N. Gradusov // Avtomatika na transporte. — 2024. — T. 10. — № 4. — S. 360–371. DOI:https://doi.org/10.20295/2412-9186- 2024-10-04-360-371.
3. Kazanskaya L. F. Bespilotnye poezda pri gruzovyh i pas- sazhirskih zheleznodorozhnyh perevozkah / L. F. Kazanskaya, P. V. Alpaeva, A. A. Schipicyna // Transportnoe delo Rossii. — 2023. — № 6. — S. 128–131.
4. Rozenberg E. N. Kompleksnye resheniya po povysheniyu propusknoy sposobnosti zheleznyh dorog / E. N. Rozen- berg, I. N. Rozenberg, A. V. Ozerov // Trudy AO «NIIAS». — 2021. — T. 1. — Vyp. 11. — S. 32–47.
5. Alekseev A. S. Model' ocenki pokazateley bezopasno- sti funkcionirovaniya avtomatizirovannyh sistem dispetcherskogo upravleniya dvizheniem poezdov / A. S. Alekseev, A. V. Gorelik, V. Yu. Gorelik, A. V. Istomin i dr. // Avtomatizaciya v promyshlennosti. — 2025. — № 1. — S. 10–13.
6. Fedorov V. A. Svertochnye neyronnye seti kak instru- ment obnaruzheniya ob'ektov zheleznodorozhnoy infra- struktury / V. A. Fedorov, O. M. Ogorodnikova // Avtomatizaciya v promyshlennosti. — 2024. — № 12. — S. 20–23. — DOI:https://doi.org/10.25728/avtprom.2024.12.04.
7. Fedorov V. A. AI-based train localization using railway infra- structure object detection / V. A. Fedorov // Transportation Development Research. — 2024. — Vol. 2. — Iss. 2. — Pp. 1–11. — DOI:https://doi.org/10.55121/tdr.v2i2.379.
8. Faghih-Roohi S. Deep convolutional neural networks for detection of rail surface defects / S. Faghih-Roohi, S. Hajizadeh, A. Nunez // 2016 International joint confer- ence on neural networks (IJCNN). IEEE, July 2016. — Pp. 2584–2589. — DOI:https://doi.org/10.1109/IJCNN.2016.7727522.
9. Zhao Y. A review on rail defect detection systems based on wireless sensors / Y. Zhao, Z. Liu, D. Yi // Sensors. — 2022. — Vol. 22. — Iss. 17. — P. 6409. — DOI:https://doi.org/10.3390/s22176409.
10. Wei X. Railway track fastener defect detection based on image processing and deep learning techniques: A com- parative study / X. Wei, Z. Yang, Y. Liu // Engineering Applications of Artificial Intelligence. — 2019. — Vol. 80. — Pp. 66–81. — DOI:https://doi.org/10.1016/j.engappai.2019.01.008.
11. Wu Y. Automatic detection of arbitrarily oriented fastener defect in high-speed railway / Y. Wu, Y. Qin, Y. Qian // Automation in Construction. — 2021. — Vol. 131. — DOI:https://doi.org/10.1016/j.autcon.2021.103913.
12. Sychugov A. N. Primenenie neyronnyh setey dlya ras- poznavaniya ob'ektov na zheleznodorozhnom transpor- te / A. N. Sychugov, V. N. Miheychikov, M. V. chernyshov // Izvestiya Peterburgskogo universiteta putey soobsche- niya. — 2023. — T. 20. — Vyp. 2. — S. 478–491. — DOI:https://doi.org/10.20295/1815-588X2023-2-478-491.
13. Fedorov V. A. Railway infrastructure instance segmentation based on convolutional neural networks / V. A. Fedorov // 2023 International Russian Automation Conference (RusAutoCon), IEEE, September 2023. — Pp. 443–447. — DOI:https://doi.org/10.1109/RusAutoCon58002.2023.10272908.
14. Ogorodnikova O. M., Method of ripe tomato detecting for a harvesting robot / O. M. Ogorodnikova, W. Ali // AIP Conference Proceedings, May 2019. — DOI:https://doi.org/10.1063/1.5134297.
15. Everingham M. The Pascal visual object classes (VOC) chal- lenge / M. Everingham, L. Van Gool, C. K. Williams // International journal of computer vision. — 2010. — Vol. 88. — Pp. 303–338. — DOI:https://doi.org/10.1007/s11263-009-0275-4.
