Refereed article in conference proceedings (A4)

Extended Panorama Tracking Algorithm for Augmenting Virtual 3D Objects In Outdoor Environments




List of Authors: Mikko Forsman, Jukka Arvo, Teijo Lehtonen

Conference name: Proceedings of International Conference on Virtual Systems and Multimedia

Publication year: 2016

Book title *: Proceedings of the 2016 international conference on virtual systems and multimedia (VSMM)

Number of pages: 8

ISBN: 978-1-4673-8993-8

ISSN: 2474-1485

DOI: http://dx.doi.org/10.1109/VSMM.2016.7863190


Abstract

Augmented Reality (AR) is currently gaining popularity in multiple professional use cases. However, applying AR for outdoor environments is still very difficult, especially when targeting customer-grade mobile devices. This paper presents extensions to a previous panoramic tracking algorithm, which is a visual tracking method for unprepared outdoor use cases. The panorama tracker is extended by Unity3D game engine integration, and by implementing two distinct initialization systems based on Global Positioning System (GPS) for the tracker. The enhanced panorama tracker is benchmarked with 30 test videos, and the test results indicate that the accuracy of the tracker is sufficient for visualization tasks in outdoor environments. In addition, performance numbers show that the implemented tracker can often reach real-time performance on some consumer-grade tablets. Accuracies of the new initialization systems seem to be sufficient, as long as previously known accurate GPS locations are used instead of internal sensors from the mobile devices.


Last updated on 2021-24-06 at 09:23