TRAFFIC SIGN AND LIGHT DETECTION USING YOLOV5

Abstract

Detecting traffic signs and lights has been a problem for intelligent vehicles for a long time.Prior to classifying traffic signs and lights, this provides an efficient method for managing the inventory

of traffic signs and lights for driver assistance or autonomous vehicles. Object identification mod- els like Fast RCNN and Faster RCNN have been applied to this issue. These approaches’ main

drawbacks are their slowness and inability to sort in real time. Deep learning was used to create the Convolutional Neural Network (CNN) for visual object detection. This has helped to fix the problems with traditional object recognition. In this proposed plan, CNN and YOLO architecture have been chosen as ways to find and sort things. Here, the newest version of the YOLOV5 was used. The YOLOV5 is mostly used because it is fast, has a simple design, and works well for most object detection tasks. In this system, a method for detecting and identifying traffic signs and lights is proposed. This method, along with deep CNN, is used to classify traffic signs and lights. The proposed system works in real time to find and identify images of traffic lights and signs.