A Distributed and Collaborative vSLAM Framework for Real-Time Localisation in Huge Environments for Mobile Devices

Abstract

This poster proposes a collaborative vSLAM system for mobile Augmented Reality applications in huge environments. The system combines the efficiency of local vSLAM methods with the accuracy of large-scale mapping techniques. In order to overcome the efficiency problems of large-scale algorithms,a distributed architecture in the Cloud that keeps the consistency of the map is proposed. To achieve it, the data coming from multi-user local vSLAM algorithms is fused with the large-scale map stored in the Cloud.