Design and Implementation of an Automatic Vehicle License Recognition System for End-of-Life Vehicle Dismantling Scenarios

Yining Liu

doi:10.54097/s0xs3m35

Authors

Yining Liu

DOI:

https://doi.org/10.54097/s0xs3m35

Keywords:

End-of-life vehicle license recognition, YOLOv8, PaddleOCR, post-processing rule engine, document digitization

Abstract

The accuracy and efficiency of vehicle license information entry directly determines whether such operations can truly implement the Standards for Motor Vehicle Recycling and Dismantling Operations and further advance the digitalization of compliant supervision in this sector. To address challenges in this scenario, such as severely damaged documents, complex shooting conditions, and strong field logic, this paper designs and implements an automatic vehicle license information recognition system for end-of-life vehicle dismantling enterprises. The system first employs a YOLOv8 model combined with a multi-frame stability criterion to achieve robust detection of the vehicle license region and capture clear images under complex shooting environments. Subsequently, PaddleOCR is utilized for text extraction. Finally, a multi-layer post-processing engine integrating regular expression matching, semantic analysis, and business rules corrects and structures the raw OCR results, outputting standardized vehicle information. Experiments demonstrate that the designed post-processing rule engine effectively corrects the majority of OCR recognition errors. Ultimately, the system is integrated into a desktop application featuring data visualization and browser automation form-filling capabilities, achieving an end-to-end automated process from physical document scanning to business system entry, thereby enhancing operational efficiency and data accuracy.

Downloads

Download data is not yet available.

References

[1] Du, Y., Li, C., Guo, R., et al. (2020). PP-OCR: A practical ultra lightweight OCR system. arXiv. https://doi.org/10.48550/arXiv.2009.09941

[2] Huang, Z., Chen, K., He, J., et al. (2020). LayoutLM: Pre-training of text and layout for document image understanding. In Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining.

[3] Redmon, J., Divvala, S., Girshick, R., et al. (2016). You only look once: Unified, real-time object detection. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.

[4] Ministry of Commerce of the People's Republic of China. (2020). Implementation rules for the administrative measures on the recycling of end-of-life vehicles. (in Chinese)

[5] Liu, W., & Zhang, Q. (2021). Research on vehicle license recognition system based on deep learning. Computer Engineering and Design, 42(5), 1342–1348. (in Chinese)

[6] Du, Y., Li, C., Guo, R., et al. (2021). PP-OCRv2: Indispensable strategies for making training better and faster. arXiv. https://doi.org/10.48550/arXiv.2109.03144

[7] Wang, H. (2022). Discussion on informatization construction of end-of-life vehicle recycling and dismantling industry. Renewable Resources and Circular Economy, 15(3), 12–15. (in Chinese)

[8] Chen, G. (2023). Computer vision: Fundamental algorithms and engineering practices. Publishing House of Electronics Industry. (in Chinese)

[9] Ge, Z., Liu, S., Wang, F., et al. (2021). YOLOX: Exceeding YOLO series in 2021. arXiv. https://doi.org/10.48550/arXiv.2107.08430