Use of a Deep Neural Network for Object Detection in a Mobile Robot Application

Abstract

This paper presents the implementation of a deep neural network for object detection in a mobile robot application. A mobile robot was designed and developed to work in dangerous environments. The mobile robot will move either tele-operated through the image transmitted by the video camera to a tablet or computer, or it will move autonomously in the environment in which it operates. In autonomous mode the mobile robot must recognize certain objects or people in the operating environment. For this reason, we implemented with good results an object detection algorithm based on deep neural network.