Штучний інтелект

Анотація: This paper presents a methodology for automated licence plate recognition and vehicle identification designed for Ukrainian standards. The approach combines YOLOv11 object detection with EasyOCR text recognition, incorporat-ing deskewing and adaptive thresholding preprocessing techniques. A custom dataset of 467 images containing Ukrainian DSTU-compliant licence plates was created from the Auto. RIA platform. The system employs format-specific error correction and character substitution rules. Experimental results demonstrate an F₁ score of 0.92 at IoU@0.5 for detection tasks, with performance variations across vehicle classes revealing challenges in commercial vehicle recognition and motorcycle licence plate processing. This work contributes annotated dataset for Ukrainian licence plate recognition and provides a practical foundation for automated traffic monitoring systems.

Ключові слова: Deep learning, YOLO, EasyOCR, deskewing, computer vision, recognition, identification

Посилання:

1. Sanzharov, D. R., & Filimonova, T. O. (2025). Identifikatsiia transportnykh zaso- biv ta nomernykh znakiv z vykorystanniam modeli simeistva Ultralitics YOLO. Infor- matsiini Tekhnolohii Ta Kiberbezpeka V Umovakh Voiennoho Chasu, 6–9.
2. Moussaoui, H., Akkad, N. E., Benslimane, M., El-Shafai, W., Baihan, A., Hewage, C., & Rathore, R. S. (2024). Enhancing automated vehicle identification by inte- grating YOLO v8 and OCR techniques for high-precision license plate detection and recognition. Scientific Reports, 14(1), 14389. https://www.nature.com/articles/s41598-024-65272-1.pdf
3. Dong, M., He, D., Luo, C., Liu, D. & Zeng, W. (2017, ). A CNN-Based Approach for Automatic License Plate Recognition in the Wild. BMVC. https://bmva-archive.org.uk/bmvc/2017/papers/paper175/paper175. pdf
4. Laroca, R., Severo, E., Zanlorensi, L. A., Oliveira, L. S., Gonçalves, G. R., Schwartz, W. R., & Menotti, D. (2018). A robust real-time automatic li-cense plate recognition based on the YOLO detector. 2018 International Joint Conference on Neural Net-works (Ijcnn), 1–10. https:// arxiv.org/pdf/1802.09567
5. Ultralitics. Ultralitics YOLO11 - Ultralitics YOLO Docs. https://docs.ultralytics.com/ models/yolo11/
6. Mao, M., & Hong, M. (2025). YOLO object de-tection for real-time fabric defect inspection in the textile industry: A review of YOLOv1 to YOLOv11. Sensors (Basel, Switzerland), 25(7), 2270. https://doi.org/ 10.3390/s25072270
7. Jaided AI. (2024). EasyOCR: Ready-to-use OCR with 80+ supported languages and all common writ-ing scripts. https:// github.com/JaidedAI/EasyOCR
8. Smelyakov, K., Chupryna, A., Darahan, D., & Midina, S. (2021). Effectiveness of Modern Text Recognition Solutions and Tools for Common Data Sources. COLINS, 154–165. https://ceur-ws.org/Vol-2870/paper15.pdf
9. OpenCV team. (2024, ). Open Source Computer Vision Library. https://github. com/opencv/opencv
10. Mauder, M. (2024, ). deskew: Skew cor- rection of images, especially documents. https://github.com/galfar/deskew
11. Ministry of Economic Development and Trade of Ukraine. DSTU 4278:2019. Road Transport. Vehicle Registration Plates. General Requirements. Rules of Applica- tion.
12. Python Software Foundation. (2024). The Py-thon Language Reference Man- ual (3.12 ed.). https://docs.python.org/3.12/reference//
13. AUTO.RIA. Automarket №1, buy and sell cars easily!. https://auto.ria.com/uk/