PO-MSCKF: An Efficient Visual-Inertial Odometry by Reconstructing the Multi-State Constrained Kalman Filter with the Pose-only Theory

Du Xueyu,Zhang Lilian,Liu Ruochen,Wang Maosong,Wu Wenqi,Mao Jun

2024-07-02

Abstract:Efficient Visual-Inertial Odometry (VIO) is crucial for payload-constrained robots. Though modern optimization-based algorithms have achieved superior accuracy, the MSCKF-based VIO algorithms are still widely demanded for their efficient and consistent performance. As MSCKF is built upon the conventional multi-view geometry, the measured residuals are not only related to the state errors but also related to the feature position errors. To apply EKF fusion, a projection process is required to remove the feature position error from the observation model, which can lead to model and accuracy degradation. To obtain an efficient visual-inertial fusion model, while also preserving the model consistency, we propose to reconstruct the MSCKF VIO with the novel Pose-Only (PO) multi-view geometry description. In the newly constructed filter, we have modeled PO reprojection residuals, which are solely related to the motion states and thus overcome the requirements of space projection. Moreover, the new filter does not require any feature position information, which removes the computational cost and linearization errors brought in by the 3D reconstruction procedure. We have conducted comprehensive experiments on multiple datasets, where the proposed method has shown accuracy improvements and consistent performance in challenging sequences.

Robotics,Computer Vision and Pattern Recognition

What problem does this paper attempt to address?

The paper attempts to address the problem of how to improve accuracy and robustness in Visual-Inertial Odometry (VIO) systems while maintaining efficiency and model consistency. Specifically, in traditional multi-view geometric descriptions, the reprojection error is related to the camera motion and the 3D positions of feature points, which leads to most existing VIO algorithms requiring simultaneous pose estimation and 3D reconstruction. As the number of observed features increases, the number of optimization parameters increases sharply, resulting in decreased efficiency. Additionally, the traditional Multi-State Constraint Kalman Filter (MSCKF) removes 3D point residuals through spatial projection operations, which avoids 3D reconstruction but affects model consistency and introduces linearization errors due to inaccurate 3D estimates.

To address these issues, the paper proposes a new Multi-State Constraint Kalman Filter based on Pose-Only (PO) theory (PO-MSCKF). The PO theory allows the reprojection error to be obtained solely through the camera pose, thereby avoiding the need for 3D feature position information and eliminating the computational cost and linearization errors in the 3D reconstruction process. Through this method, PO-MSCKF not only maintains the concise state space of MSCKF but also improves model consistency and accuracy. Experimental results show that PO-MSCKF performs better on various datasets, especially in challenging sequences.