English

Revisiting visual-inertial structure from motion for odometry and SLAM initialization

Computer Vision and Pattern Recognition 2021-01-29 v2 Robotics

Abstract

In this paper, an efficient closed-form solution for the state initialization in visual-inertial odometry (VIO) and simultaneous localization and mapping (SLAM) is presented. Unlike the state-of-the-art, we do not derive linear equations from triangulating pairs of point observations. Instead, we build on a direct triangulation of the unknown 3D3D point paired with each of its observations. We show and validate the high impact of such a simple difference. The resulting linear system has a simpler structure and the solution through analytic elimination only requires solving a 6×66\times 6 linear system (or 9×99 \times 9 when accelerometer bias is included). In addition, all the observations of every scene point are jointly related, thereby leading to a less biased and more robust solution. The proposed formulation attains up to 5050 percent decreased velocity and point reconstruction error compared to the standard closed-form solver, while it is 4×4\times faster for a 77-frame set. Apart from the inherent efficiency, fewer iterations are needed by any further non-linear refinement thanks to better parameter initialization. In this context, we provide the analytic Jacobians for a non-linear optimizer that optionally refines the initial parameters. The superior performance of the proposed solver is established by quantitative comparisons with the state-of-the-art solver.

Keywords

Cite

@article{arxiv.2006.06017,
  title  = {Revisiting visual-inertial structure from motion for odometry and SLAM initialization},
  author = {Georgios Evangelidis and Branislav Micusik},
  journal= {arXiv preprint arXiv:2006.06017},
  year   = {2021}
}
R2 v1 2026-06-23T16:13:02.221Z