Neural controller of autonomous driving mobile robot by an embedded camera

Abstract

The purpose of this paper is to build an autonomous RC Car that uses Artificial Neural Network (ANN) for control. It describes the theory behind the neural network and autonomous vehicles, and how a prototype with a camera as its only input can be designed to test and evaluate the algorithm capabilities. The ANN is a good algorithm that could help recognize patterns in an image, it can with a training set, containing 2000 images, classify an image with 96% of accuracy rate. The main contribution of this paper consists in using a single camera for navigation, possibly for obstacle avoidance.