Paper Title
Automatic Road Sign Detection and Recognition using CNN

Abstract
Automatic detection of road signs plays an important role in managing the road safety. It provides precise way to manage road signs. There are numerous amounts of approaches are present to detect the traffic sign and perform really well. These approaches are helpful to advance driver assistance system and autonomous vehicle. TSR has been analyzed for several years but mainly focus of existing work is symbol based road signs. In this paper, we use a Convolutional neural network (CNN) approach to deal with the detection and recognition of traffic signs. CNN is applied to detection of various traffic sign categories which are present in our dataset. Results are tested on highly challenging traffic sign images which are present in our dataset. This paper proposes a system to acknowledge various traffic signs, in video sequences captured by a camera placed on a car. This system mainly consists of two stages first one is detection and other one is recognition. In the first phase, pre- processing, enhancement and segmentation is performed on the traffic sign images. In the second phase output of first phase is given as an input to test against some set of particular features to check whether it belongs to the group of traffic signs or not and based on these properties they are categorized into various groups. Our model is validated on the German traffic sign dataset and higher efficiency is achieved. Keywords - Traffic Signs, Machine Learning Algorithm, Traffic Sign Detection and Recognition (TSDR), Convolutional Neural Network (CNN)

Author - Chitra Kumari, Yogesh Kakde, Astha Verma, Monika

Published : Volume-8,Issue-5  ( May, 2021 )

	|		PDF	|	Viewed - 44	|	Published on 2021-09-15