Tightly-coupled Magneto-Visual-Inertial Fusion for Long Term Localization in Indoor Environment

Jade Coulin, Richard Guillemard, Vincent Gay-Bellile, Cyril JOLY, Arnaud de La Fortelle

Abstract:

We propose in this paper a tightly-coupled fusion of
visual, inertial and magnetic data for long-term localization in in-
door environment. Unlike state-of-the-art Visual-Inertial SLAM
(VISLAM) solutions that reuse visual map to prevent drift, we
present in this paper an extension of the Multi-State Constraint
Kalman Filter (MSCKF) that takes advantage of a magnetic
map. It makes our solution more robust to variations of the
environment appearance. The experimental results demonstrate
that the localization accuracy of the proposed approach is almost
the same over time periods longer than a year.