Publication Date: 2022/10/07
Abstract: Traffic signs are a crucial part of our road environment. They provide crucial information, sometimes compelling recommendations, to ensure that driving behaviors are adjusted and that any currently enforced traffic regulations are observed. With majority of modern automobiles equipped with an automated driving assistance systems a robust and efficient traffic sign classifier would be considered a must. We propose a Traffic Sign Recognition system which follows a neural network-based approach that uses YOLOv3 (You Only Look Once Version 3) as object detector rather than a classifier followed by a CNN (Convolutional Neural Network) to classify traffic signs. This approach of dividing the modules to compute single task turns out to improve the system’s performance even with limited training thus providing a better platform for development of models to solve similar tasks.
Keywords: YOLOv3, CNN, Traffic Sign, Image, Classification, Detection, Recognition.
DOI: https://doi.org/10.5281/zenodo.7155146
PDF: https://ijirst.demo4.arinfotech.co/assets/upload/files/IJISRT22SEP574.pdf
REFERENCES