An Omnidirectional System for Navigation in Virtual Environments

Abstract

Virtual Reality (VR) is one of the newest technological domains with revolutionary applicability for the tomorrow's Future Internet, including visions of the Internet of Things. The sensation of total immersion in Virtual Environment (VE) is still unresolved. Therefore, our work proposes a new omnidirectional locomotion interface for navigation in VEs. The novel interface was built from an ordinary unidirectional treadmill, a new mechanical device, a motion capturing system to track the human walking and a control method using artificial intelligence techniques. A neural network is used to predict the motion of the new interface based on user's body motion and information about VE. The feasibility of the proposed system is verified through experiments and the preliminary results suggest that the new interface performs very well in a simplest VE based on our control method.