Digital Twin and Blockchain for Smart Cities

18Deep Learning-Based Traffic Sign Detection and Recognition for Autonomous Vehicles

Murali Krishnan Mani¹, Sonaa Rajagopal¹*, Kavitha D.¹ and Saravanabalagi Ramachandran²

¹School of Computer Science and Engineering, Vellore Institute of Technology, Chennai, Tamil Nadu, India

²Computer Vision for Autonomous Vehicles in Visual Place, Recognition and Topological SLAM, Maynooth University, Maynooth, County Kildare, Ireland

Abstract

This study aims to create a model for traffic sign detection and recognition to categorize traffic signs in real-world scenarios with precision using pre-existing YOLOv9 model. The project includes numerous important stages to accomplish this goal, such as thorough dataset collecting, model training, and optimization procedures. Through lengthy training on large-scale datasets with a variety of traffic sign images, the YOLOv9-based model learns complex patterns and features indicative of different traffic signs. To further strengthen the system’s resistance to problems like changing lighting, occlusions, and distortions frequently seen in real-world situations, sophisticated image processing techniques are used. The system aims to attain superior performance in traffic sign detection and recognition tasks by means of repeated refining and optimization. With regard to practical applications, the resulting approach has a great deal of promise, especially for ADAS and driverless cars. Through a smooth integration into existing systems, the suggested approach ...