A Review on Road Sign Identification Using Machine Learning Techniques

The evaluation of current Traffic Sign Recognition (TSR) methodologies shows that most systems operate as simple classifiers without translating visual data into actionable driving commands. This project successfully addresses this gap by developing an end-to-end perception-to-action pipeline that utilizes a Convolutional Neural Network trained on an Indian dataset of 10 high-priority sign categories. The system architecture includes a unique Command Logic Layer that maps identified signage to specific vehicle actions, such as braking or rerouting, under a strict 90% confidence gate. Experimental results demonstrate that the system maintains a target accuracy exceeding 95% while processing a live vision feed from a laptop-connected vision sensor. By integrating a Live Dashboard that provides immediate visual alerts and warnings, the system offers a robust safety solution for Advanced Driver Assistance Systems. Overall, the developed system provides a cost-effective and computationally efficient approach to enhancing road safety in India, ensuring that critical traffic information is not only recognized but acted upon in real time to prevent accidents.

[1]   M. Rahman, P. T. Mazumder, M. A. S. Abid, A. H. Siam, and M. S. Islam, "BanglaRoadSignNet: A YOLO-Based Framework for Efficient Road Sign Detection and Recognition in Bangladesh," in 2025 IEEE, Jan. 2025.   [2]  Ms. R. Dhivyapraba, Kaushik P, Dr. K. Kavitha, and Karthik P, "AI enhanced Autonomous Traffic Sign Recognition and Action System," in 2025 IEEE, Jan. 2025.   [3]   M. Abdelazeem, M. Ghantous, and H. Soubra, "Vision-Based Real-time Hazard Identification and Alerting for Rider Assistance," Media Engineering Technology, The German University In Cairo, 2024.   [4]  A. Sriram.D, Radhakrishnan.P, Raja.E, and Praveen.M, "A Traffic Road Sign Recognition System using Raspberry pi with Open CV Libraries Based On Embedded Technologies," in 2023 International Conference on Computer Communication and Informatics (ICCCI), Jan. 2023.   [5]   J. Wang, Y. Chen, Z. Dong, and M. Gao, "Improved YOLOv5 network for real-time multi-scale traffic sign detection," Neural Computing and Applications, vol. 35, pp. 7853–7865, 2023.    [6]   M. Tamilselvi, Iyswariya A, K. Naresh Kumar Thapa, B. Anni Princy, Vinithra Banu.T, and V. Samuthira Pandi, "Design and Development of Artificial Intelligence based Real-Time Traffic Sign Identification Scheme using Novel Learning Strategy," 2023.   [7]   J. Chaudhari, R. Patel, and D. K. Patel, "Augmented YOLOv5 for Indian Traffic Signs Identification," Sankalchand Patel College of Engineering, 2023.   [8]   R. Tarun and B. Priya Esther, "Real-time regional road sign detection and identification using Raspberry Pi," SRM Institute of Science and Technology, 2023.   [9]   T. SubhaMastan Rao, S. Anjali Devi, B. Jhansi Vazram, and B. Srinivasa Rao, "A Novel Approach for Detecting Traffic Signs using Deep Learning," CMR Technical Campus, 2023.   [10]  N. Bhatt, P. Laldas, and V. B. Lobo, "A Real-Time Traffic Sign Detection and Recognition System on Hybrid Dataset using CNN," St. John College of Engineering and Management, 2022.