Intelligent Traffic Sign Detection Using Yolov9
DOI:
https://doi.org/10.47701/icohetech.v5i1.4205Keywords:
Traffic Sign Detection, YOLOv9, Computer Vision, Artificial IntelligenceAbstract
This research examines the automatic detection and classification of traffic signs using artificial intelligence (AI) and computer vision technologies. As urban traffic increases, quickly and accurately recognizing traffic signs becomes a challenge, especially under adverse conditions such as bad weather and limited visibility. Conventional technologies that rely on human vision are prone to errors, so an automated solution is needed. This research uses the YOLOv9 algorithm for real-time traffic sign detection, utilizing the Generalized ELAN (GELAN) architecture that combines the advantages of CSPNet and ELAN for efficiency and accuracy. The dataset used consists of 1924 images processed through various stages, including data augmentation and normalization. The model was trained for 15 epochs with fairly high accuracy results in the prohibitory, danger, and mandatory sign categories. However, there were still some misclassifications, especially in the prohibitory category which was sometimes mistakenly detected as another category or background. Overall, the model performed well in detecting traffic signs in various environmental conditions, but still needs improvement to increase accuracy in certain cases.
References
Bochkovskiy, A., Wang, C. Y., & Liao, H. Y. M. (2020). YOLOv4: Optimal speed and accuracy of object detection. arXiv preprint, arXiv:2004.10934. https://arxiv.org/abs/2004.10934
Duan, L., Zhao, Y., Liu, J., & Zhu, Z. (2020). Robust traffic sign recognition using multi-feature fusion and a CNN. Journal of Visual Communication and Image Representation, 73, 102956. https://doi.org/10.1016/j.jvcir.2020.102956
Khalid, S., Nawaz, S., & Ullah, Z. (2020). Automated traffic sign detection and classification system using deep learning. Multimedia Tools and Applications, 79(25), 17735-17750. https://doi.org/10.1007/s11042-020-08991-1
Larsson, E., Åström, K., & Heyden, A. (2019). Real-time traffic sign detection using an extended AdaBoost algorithm. Computer Vision and Image Understanding, 184, 58-67. https://doi.org/10.1016/j.cviu.2019.05.002
Redmon, J., Divvala, S., Girshick, R., & Farhadi, A. (2016). You only look once: Unified, real-time object detection. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (pp. 779-788). https://doi.org/10.1109/CVPR.2016.91
Sermanet, P., & LeCun, Y. (2011). Traffic sign recognition with multi-scale convolutional networks. In Proceedings of the 2011 International Joint Conference on Neural Networks (pp. 2809-2813). IEEE. https://doi.org/10.1109/IJCNN.2011.6033589
Sichkar, V. (2020). Traffic signs dataset in YOLO format [Dataset]. Benchmark. http://benchmark.ini.rub.de/?section=gtsdb&subsection=dataset
Wang, C. Y., Yeh, I. H., & Liao, H. Y. M. (2024). YOLOv9: Learning what you want to learn using programmable gradient information. arXiv preprint, arXiv:2402.13616. https://arxiv.org/abs/2402.13616
Zhou, W., Zhang, Y., Yang, F., & Wang, Q. (2019). Traffic sign recognition in the wild: A new benchmark. IEEE Transactions on Intelligent Transportation Systems, 21(5), 1900-1912. https://doi.org/10.1109/TITS.2019.2928530
