A Parallax Based Robust Image Matching for Improving Multisensor Navigation in GNSS-denied Environments

Eduard Angelats, Pere Molina, M. Eulalia Parés, Ismael Colomina

Abstract:	In this paper we concentrate on inertial visual-aided navigation and its robustness. We present a method to remove outliers in photogrammetric measurements extracted from overlapping images, using relative orientation derived from inertial-based trajectory. By means of relative orientation, the inter-distance between rays traced from corresponding tie points, named in this paper as parallax, is used for outlier detection and isolation. Once outliers are removed, tie point measurements are used to improve the trajectory estimation. Image and object space coordinates of the corresponding tie points contribute, through appropriate modelling, as position and attitude updates to the filtering step. The proposed approach has been evaluated using real and simulated data from a mobile mapping campaign over an urban area with long GNSS outages periods. The proposed approach shows that outlier removal is feasible and the trajectory is recovered, even in presents of GNSS outages.
Published in:	Proceedings of the 27th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2014) September 8 - 12, 2014 Tampa Convention Center Tampa, Florida
Pages:	2132 - 2138
Cite this article:	Angelats, Eduard, Molina, Pere, Parés, M. Eulalia, Colomina, Ismael, "A Parallax Based Robust Image Matching for Improving Multisensor Navigation in GNSS-denied Environments," Proceedings of the 27th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2014), Tampa, Florida, September 2014, pp. 2132-2138.
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In