Polarized light-aided visual-inertial navigation system: global heading measurements and graph optimization-based multi-sensor fusion

Linlin Xia,Ruimin Liu,Daochang Zhang,Jingjing Zhang
DOI: https://doi.org/10.1088/1361-6501/ac4637
IF: 2.398
2022-02-17
Measurement Science and Technology
Abstract:Abstract Polarized skylight is as fundamental a constituent of passive navigation as the geomagnetic field. With regard to its applicability to outdoor robot localization, a polarized light-aided visual-inertial navigation system (VINS) modelization dedicated to globally optimized pose estimation and heading correction is constructed. The combined system follows typical visual simultaneous localization and mapping (SLAM) frameworks, and we propose a methodology to fuse global heading measurements with visual and inertial information in a graph optimization-based estimator. With ideas of‘adding new attributes of graph vertices and creating heading error-encoded constraint edges’, the heading, as the absolute orientation reference, is estimated by the Berry polarization model and continuously updated in a graph structure. The formulized graph optimization process for multi-sensor fusion is simultaneously provided. In terms of campus road experiments on the Bulldog-CX robot platform, the results are compared against purely stereo camera-dependent and VINS Fusion frameworks, revealing that our design is substantially more accurate than others with both locally and globally consistent position and attitude estimates. As a passive and tightly coupled navigation mode, the polarized light-aided VINS can therefore be considered as a tool candidate for a class of visual SLAM-based multi-sensor fusion.
engineering, multidisciplinary,instruments & instrumentation
What problem does this paper attempt to address?