Automatic Detection of Navigational Signs on Inland Waterways Using YOLO Neural Networks

ABSTRACT: The study analysed the detection of navigation signs for inland navigation using YOLO neural networks. All major versions of the network available at the time of the study were analysed, i.e. from YOLO 1 to YOLO 12. The study considered two criteria: detection efficiency and detection accuracy. The first case is related to applications requiring the highest possible number of object detections, while the second is related to mapping tasks, where the accuracy of determining the location of a sign plays an important role. The results of the study showed that different efficiencies can be expected for the neural models studied. The latest models do not always prove to be the best. In terms of detection efficiency, the YOLO 4 network proved to be the best model, while in terms of sign detection precision, YOLO v7 had the highest horizontal accuracy and YOLO v10 had the highest vertical accuracy. The results of the study indicated that in some cases, attention should be paid to the oldest versions of YOLO, such as YOLO V1 and V2, which significantly reduce false detections.

KEYWORDS:

REFERENCES

Azimi, S.; Salokannel, J.; Lafond, S.; Lilius, J.; Salokorpi, M.; Porres, I. A Survey of Machine Learning Approaches for Surface Maritime Navigation. 2020. Available online: http://hdl.handle.net/2117/329714 (accessed on 12 February 2025).

Donandt, K.; Böttger, K.; Söffker, D. Short-Term Inland Vessel Trajectory Prediction with Encoder-Decoder Models. In Proceedings of the 2022 IEEE 25th International Conference on Intelligent Transportation Systems (ITSC), Macau, China, 8–12 October 2022; pp. 974–979. https://doi.org/10.1109/ITSC55140.2022.9922148.

Agorku, G.; Hernandez, S.; Falquez, M.; Poddar, S.; Amankwah-Nkyi, K. Traffic Cameras to Detect Inland Waterway Barge Traffic: An Application of Machine Learning, Computer Vision and Pattern Recognition. arXiv 2024, arXiv:2401.03070.

Hart, F.; Okhrin, O.; Treiber, M. Vessel-Following Model for Inland Waterways Based on Deep Reinforcement Learning. Ocean. Eng. 2023, 281, 114679. https://doi.org/10.1016/j.oceaneng.2023.114679.

Vanneste, A.; Vanneste, S.; Vasseur, O.; Janssens, R.; Billast, M.; Anwar, A.; Mets, K.; De Schepper, T.; Mercelis, S.; Hellinckx, P. Safety Aware Autonomous Path Planning Using Model Predictive Reinforcement Learning for Inland Waterways. In Proceedings of the IECON 2022–48th Annual Conference of the IEEE Industrial Electronics Society, Brussels, Belgium, 17–20 October 2022; pp. 1–6. https://doi.org/10.1109/IECON49645.2022.9968678.

Qiao, Y.; Yin, J.; Wang, W.; Duarte, F.; Yang, J.; Ratti, C. Survey of Deep Learning for Autonomous Surface Vehicles in Marine Environments. IEEE Trans. Intell. Transp. Syst. 2023, 24, 3678–3701. https://doi.org/10.1109/TITS.2023.3235911.

Hao, G.; Xiao, W.; Huang, L.; Chen, J.; Zhang, K.; Chen, Y. The Analysis of Intelligent Functions Required for Inland Ships. J. Mar. Sci. Eng. 2024, 12, 836. https://doi.org/10.3390/jmse12050836.

Li, Y.; Hu, Y.; Rigo, P.; Lefler, F.E.; Zhao, G.; Eds. Proceedings of PIANC Smart Rivers 2022: Green Waterways and Sustainable Navigations; Lecture Notes in Civil Engineering; Springer: Singapore, 2023; Volume 264. https://doi.org/10.1007/978-981-19-6138-0.

Fan, W.; Zhong, Z.; Wang, J.; Xia, Y.; Wu, H.; Wu, Q.; Liu, B. Vessel-Bridge Collisions: Accidents, Analysis, and Protection. China Journal of Highway and Transport. 2024, 37(5), 38–66.

Łubczonek, J., i M. Włodarczyk. Wykorzystanie geobazy danych w procesie tworzenia elektronicznych map nawigacyjnych dla żeglugi śródlądowe [Application of geodatabase in the process of creation electronic navigational charts for inland shipping]. Archiwum Fotogrametrii, Kartografii i Teledetekcji, t. 21, 2010, s. 221–34.

Łubczonek, J. Opracowanie i implementacja elektronicznych map nawigacyjnych dla systemu RIS w Polsce [Elaboration and implementation of electronic navigational charts for RIS in Poland]. Roczniki Geomatyki 2015, 13, 359–368.

Adamski, P.; Lubczonek, J. A Comparative Analysis of the Usability of Consumer Graphics Cards for Deep Learning in the Aspects of Inland Navigational Signs Detection for Vision Systems. Appl. Sci. 2025, 15, 5142. https://doi.org/10.3390/app15095142

SIGNI. European Code for Signs and Signals on Inland Waterways: Resolution No. 90; United Nations: New York, NY, USA, 2018.

Redmon J. Darknet: Open Source Neural Networks in C. Online source: https://pjreddie.com/darknet. Access date 13.07.2025

Jocher G.; Qiu J.; Charasia A. Ultralyiics YOLO. Online source: https://github.com/ultralytics/ultralytics Access date:13.07.2025

Redmon, J., Divvala, S., Girshick, R., & Farhadi, A. You only look 1182 once: Unified, real-time object detection. In Proc. IEEE Conf. Comput (Vol. 1183, pp. 779-788).

Redmon, J., Farhadi, A. YOLO9000: Better, Faster, Stronger. Proceedings of the IEEE conference on computer vision and pattern recognition

Farhadi A., Redmon J. "Yolov3: An incremental improvement." In Computer vision and pattern recognition, vol. 1804, pp. 1-6. Berlin/Heidelberg, Germany: Springer, 2018.

Bochkovskiy, Alexey & Wang, Chien-Yao & Liao, Hong-yuan. (2020). YOLOv4: Optimal Speed and Accuracy of Object Detection. 10.48550/arXiv.2004.10934.

Wang, C., Bochkovskiy, A., & Liao, H. M. (2022). YOLOv7: Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors. arXiv (Cornell University). https://doi.org/10.48550/arxiv.2207.02696

Bochkovskiy, A. Open Source Neural Networks in C. Online source: https://github.com/AlexeyAB/darknet Access date:13.07.2025

Khanam, R., & Hussain, M. (2024). What is YOLOv5: A deep look into the internal features of the popular object detector. arXiv (Cornell University). https://doi.org/10.48550/arxiv.2407.20892

Li, C., Li, L., Jiang, H., Weng, K., Geng, Y., Li, L., Ke, Z., Li, Q., Cheng, M., Nie, W., Li, Y., Zhang, B., Liang, Y., Zhou, L., Xu, X., Chu, X., Wei, X., & Wei, X. (2022). YOLOV6: A Single-Stage Object Detection Framework for Industrial Applications. arXiv (Cornell University). https://doi.org/10.48550/arxiv.2209.02976

R. Varghese and S. M., "YOLOv8: A Novel Object Detection Algorithm with Enhanced Performance and Robustness," 2024 International Conference on Advances in Data Engineering and Intelligent Computing Systems (ADICS), Chennai, India, 2024, pp. 1-6, doi: 10.1109/ADICS58448.2024.10533619.

Wang, C., Yeh, I., & Liao, H. M. (2024). YOLOV9: Learning what you want to learn using programmable gradient information. arXiv (Cornell University). https://doi.org/10.48550/arxiv.2402.13616

Wang, A., Chen, H., Liu, L., Chen, K., Lin, Z., Han, J., & Ding, G. (2024). YOLOV10: Real-Time End-to-End Object Detection. arXiv (Cornell University). https://doi.org/10.48550/arxiv.2405.14458

Khanam, R., & Hussain, M. (2024b). YOLOV11: An overview of the key architectural enhancements. arXiv (Cornell University). https://doi.org/10.48550/arxiv.2410.17725

Tian, Y., Ye, Q., & Doermann, D. (2025). YOLOV12: Attention-Centric Real-Time Object Detectors. arXiv (Cornell University). https://doi.org/10.48550/arxiv.2502.12524

Nvidia Corporation. CUDA Toolkit Documentation 12.4 Available online: https://docs.nvidia.com/cuda/archive/12.4.0/ (accessed on 13 July 2025).

Paszke, A.; et al. PyTorch: An Imperative Style, High-Performance Deep Learning Library. arXiv 2019, arXiv:1912.01703. Available online: http://arxiv.org/abs/1912.01703 (accessed on 12 February 2025)

Citation note:

Adamski P., Łubczonek J.: Automatic Detection of Navigational Signs on Inland Waterways Using YOLO Neural Networks. TransNav, the International Journal on Marine Navigation and Safety of Sea Transportation, Vol. 19, No. 4, doi:10.12716/1001.19.04.18, pp. 1203-1208, 2025

BibTeX EndNote

File downloaded 2 times