Unleashing the Power of Discrete-Time State Representation: Ultrafast Target-Based IMU-Camera Spatial-Temporal Calibration
Junlin Song, Antoine Richard, and Miguel Olivares-Mendez
AI summary
Problem
Existing target-based IMU-camera calibration methods rely on continuous-time B-spline representations, which incur high computational costs due to high-dimensional state vectors and complex derivative calculations.
Approach
The method replaces continuous-time states with a discrete-time formulation that aggregates IMU measurements via preintegration, jointly estimates gravity, and uses higher-order midpoint integration to drastically reduce optimization dimensions.
Key results
- ~600x faster calibration than Kalibr on standard datasets
- Competitive spatial-temporal accuracy matching continuous-time baselines
- First method to jointly estimate gravity with IMU preintegration
- Zero downstream VIO accuracy loss
Why it matters
Enables rapid, scalable calibration for millions of consumer and industrial visual-inertial devices without compromising navigation or perception performance.
Abstract
Visual-inertial fusion is crucial for a large amount of intelligent and autonomous applications, such as robot navigation and augmented reality. To bootstrap and achieve optimal state estimation, the spatial-temporal displacements between IMU and cameras must be calibrated in advance. Most existing calibration methods adopt continuous-time state representation, more specifically the B-spline. Despite these methods achieve precise spatial-temporal calibration, they suf- fer from high computational cost caused by continuous-time state representation. To this end, we propose a novel and extremely efficient calibration method that unleashes the power of discrete-time state representation. Moreover, the weakness of discrete-time state representation in temporal calibration is tackled in this paper. With the increasing production of drones, cellphones and other visual-inertial platforms, if one million devices need calibration around the world, saving one minute for the calibration of each device means saving 2083 work days in total. To benefit both the research and industry communities, the open-source implementation is released at https://github.com/JunlinSong/DT-VI-Calib.