Enhanced Traffic Sign Recognition System Using Convolutional Neural Networks
A. Mary Jenifer, R. Balamanigandan
Abstract
A Safe navigation requires that the vehicle will be able to detect the road edges and obstacles while respecting the road signs. Sometimes It is a critical domain with a significant impact on driverless vehicle navigation by interpreting both permanent and temporary road signs. This field, which includes Traffic Sign Detection and Classification makes up a comprehensive recognition system critical for autonomous transportation. This research work focuses on the exploration and refinement of Traffic Sign Recognition, particularly concerning its deployment on portable devices. While prioritizing real-time responsiveness, the model's accuracy in detecting traffic signs is updated. The model developed in this study performs exceptionally well, with an accuracy of 92.5%. The experimental results validate the efficacy of the network in successfully classifying traffic signs. This research study presents a streamlined and high-performing Traffic Sign Recognition model for portable devices with exceptional accuracy, Precision, Recall, F1-Score and validation through rigorous experimentation, thereby proving its suitability for real-world deployment in unmanned driving scenarios.