Real-Time Urban Traffic Perception for Resource-Constrained Embedded Vision: A Review

Deyu Yu

doi:10.54691/z670ae37

Authors

Deyu Yu

DOI:

https://doi.org/10.54691/z670ae37

Keywords:

Intelligent Transportation Systems (ITS), real-time urban traffic perception, embedded vision, deep learning, edge computing.

Abstract

Urban mobility is essential to the economic and social vitality of cities, but congestion, pollution, and security issues continue to plague urban transport systems. Conventional traffic sensing and control techniques cannot provide timely and scalable awareness at a large scale in complex urban environments, so the development of intelligence data-driven solutions is justifiable. Recent years have witnessed great strides in deep learning-based vision for tasks such as detection, segmentation, and multi-object tracking in urban areas, but deploying such models on omnipresent but resource-constrained edge hardware is a significant bottleneck. This review highlights the potential of embedded vision and lightweight deep learning for real-time traffic understanding in ITS. It surveys model compression, pruning, and compact architectures, and edge-optimised inferences, all of which, combined with edge–cloud collaborations that consider accuracy, latency, and energy tradeoff, render the model inferencing feasible at the edge. The survey also addresses practical evaluation criteria, challenges including robustness, data drift, worst case latency, safety, as well as maintainability. Finally, it outlines a set of future trends-adaptive inference, smart offloading, hardware-aware design, and standardised benchmarks pave the way for scalable, reliable, and green ITS deployment.

Downloads

Download data is not yet available.

References

[1] Kumarage, A. S. (2004). Urban traffic congestion: The problem & solutions. Asian Econ. Rev, 2(2004), 10-19.

[2] Ravi, S., & Mamdikar, M. R. (2022, May). A review on ITS (intelligent transportation systems) technology. In 2022 International Conference on Applied Artificial Intelligence and Computing (ICAAIC) (pp. 155-159). IEEE.

[3] Aisyah, G., & Bestari, P. (2018, November). Implementation E-Tilang in Bandung to Increase Awareness of Cross as Moral Law Passed Citizenship (Civic Virtue). In Annual Civic Education Conference (ACEC 2018) (pp. 660-663). Atlantis Press.

[4] Chen, J., Wang, Q., Cheng, H. H., Peng, W., & Xu, W. (2022). A review of vision-based traffic semantic understanding in ITSs. IEEE Transactions on Intelligent Transportation Systems, 23(11), 19954-19979. IEEE.

[5] Muhammad, K., Hussain, T., Ullah, H., Del Ser, J., Rezaei, M., Kumar, N., ... & De Albuquerque, V. H. C. (2022). Vision-based semantic segmentation in scene understanding for autonomous driving: Recent achievements, challenges, and outlooks. IEEE Transactions on Intelligent Transportation Systems, 23(12), 22694-22715. IEEE.

[6] Dolatyabi, P., Regan, J., & Khodayar, M. (2025). Deep Learning for Traffic Scene Understanding: A Review. IEEE Access.

[7] Howard, A. G., Zhu, M., Chen, B., Kalenichenko, D., Wang, W., Weyand, T., ... & Adam, H. (2017). Mobilenets: Efficient convolutional neural networks for mobile vision applications. arXiv preprint arXiv:1704.04861.

[8] Zhang, S., Wen, L., Bian, X., Lei, Z., & Li, S. Z. (2018). Single-shot refinement neural network for object detection. In Proceedings of the IEEE conference on computer vision and pattern recognition (pp. 4203-4212). IEEE.

[9] Zhang, X., Zhou, X., Lin, M., & Sun, J. (2018). Shufflenet: An extremely efficient convolutional neural network for mobile devices. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (pp. 6848–6856). IEEE.

[10] Singh, R., & Gill, S. S. (2023). Edge AI: a survey. Internet of Things and Cyber-Physical Systems, 3, 71-92.

[11] Setyanto, A., Sasongko, T. B., Fikri, M. A., & Kim, I. K. (2023). Near-edge computing aware object detection: A review, 12, 2989-3011. IEEE.

[12] Beltrán-Escobar, M., Alarcón, T. E., Rumbo-Morales, J. Y., López, S., Ortiz-Torres, G., & Sorcia-Vázquez, F. D. (2024). A review on resource-constrained embedded vision systems-based tiny machine learning for robotic applications. Algorithms, 17(11), 476.

[13] Elmanaa, I., Sabri, M. A., Abouch, Y., & Aarab, A. (2023). Efficient roundabout supervision: real-time vehicle detection and tracking on Nvidia Jetson Nano. Applied sciences, 13(13), 7416.

[14] Nocua, F., Perez-Holguin, W. J., & Pardo-Beainy, C. (2025). Urban traffic monitoring based on deep learning on an embedded GPU. Expert Systems with Applications, 273, 126847.

[15] Al Amin, R., Hasan, M., Wiese, V., & Obermaisser, R. (2024). FPGA-based real-time object detection and classification system using YOLO for edge computing, 12, 73268-73278. IEEE Access.

[16] Wang, H., Mo, R., Chen, Y., Lin, W., Xu, M., & Liu, B. (2023). Pedestrian and vehicle detection based on pruning YOLOv4 with cloud-edge collaboration. Computer Modeling in Engineering & Sciences, 137(2), 1–15.

[17] El Zeinaty, C., Hamidouche, W., Herrou, G., & Menard, D. (2025). Designing Object Detection Models for TinyML: Foundations, Comparative Analysis, Challenges, and Emerging Solutions. ACM Computing Surveys.

[18] Mittal, P. (2024). A comprehensive survey of deep learning-based lightweight object detection models for edge devices. Artificial Intelligence Review, 57(9), 242.

[19] Sun, K., Wang, X., Miao, X., & Zhao, Q. (2025). A review of AI edge devices and lightweight CNN and LLM deployment. Neurocomputing, 614, 128791.

[20] Gong, T., Zhu, L., Yu, F. R., & Tang, T. (2023). Edge intelligence in intelligent transportation systems: A survey. IEEE Transactions on Intelligent Transportation Systems, 24(9), 8919-8944. IEEE.

[21] Jouini, O., Sethom, K., Namoun, A., Aljohani, N., Alanazi, M. H., & Alanazi, M. N. (2024). A survey of machine learning in edge computing: Techniques, frameworks, applications, issues, and research directions. Technologies, 12(6), 81.

[22] Xu, D., He, X., Su, T., & Wang, Z. (2023). A survey on deep neural network partition over cloud, edge and end devices. arXiv preprint arXiv:2304.10020.

[23] Ruiz-Barroso, P., Castro, F. M., & Guil, N. (2025). Real-time unsupervised video object detection on the edge. Future Generation Computer Systems, 167, 107737.

[24] Liang, S., Wu, H., Zhen, L., Hua, Q., Garg, S., Kaddoum, G., ... & Yu, K. (2022). Edge YOLO: Real-time intelligent object detection system based on edge-cloud cooperation in autonomous vehicles. IEEE Transactions on Intelligent Transportation Systems, 23(12), 25345-25360. IEEE.

[25] Humes, E., Navardi, M., & Mohsenin, T. (2023). Squeezed edge yolo: Onboard object detection on edge devices. arXiv preprint arXiv:2312.11716.

[26] Du, C., Su, S., Lin, C., Yao, Y., Jin, R., & Hong, X. (2025). A lightweight network for traffic sign detection via multiple scale context awareness and semantic information guidance. Scientific Reports, 15(1), 10110.

[27] Gu, X., & Zhang, G. (2025). PSC-YOLO: a lightweight model for urban road instance segmentation. Journal of Real-Time Image Processing, 22(2), 1-13.

[28] Li, X., Chen, C., Lou, Y. Y., Abdallah, M., Kim, K. T., & Bagchi, S. (2024). HopTrack: A Real-time Multi-Object Tracking System for Embedded Devices. arXiv preprint arXiv:2411.00608.

[29] Guerrouj, F. Z., Florez, S. A. R., El Ouardi, A., Abouzahir, M., & Ramzi, M. (2025). Quantized Object Detection for Real-Time Inference on Embedded GPU Architectures. International journal of advanced computer science and applications (IJACSA), 16(5).