An Ultrasonic Vision based on a Neural Network for Navigating Robots Through Obstacles

Abstract

The technologies used for robot navigation have their advantages and disadvantages. To improve the technology, specialists often turn to the biological world. In this article, we implemented machine vision based on ultrasound using a neural network, similar to the ultrasound vision of a bat. The ultrasound system consists of a software and hardware installation and an artificial neural network. The use of an artificial neural network increased the accuracy of positioning, as well as provided three-dimensional vision (in 3D). This article presents the learning outcomes of a fully connected neural network and the evaluation of learning outcomes.