Line Following Autonomous Driving Robot using Deep Learning

Abstract

In this paper, we proposed a deep neural network for the learning of an autonomous line following robot. In this way, the robot can autonomously drive and follow the line painted on the ground. First of all, we collected a sufficient amount of data for training the network with images and videos. Consequently, CNN and RNN models were applied to learn and classify the robot states. Experimental results in a simulator show the efficiency of the proposed approach.