Abstract:Many navigation problems can be formulated as observer design on linear observed systems with a two-frame group structure, on which an invariant filter can be implemented with guaranteed consistency and stability. It's still unclear how this could be generalized to simultaneous estimation of the poses of multiple frames and the general forms of the linear observed systems involving multiple frames remain unknown. In this letter, we propose a multi-frame group structure by semi-direct product using the two-frame group as building blocks, covering all natural extensions. More importantly, we give a systematic direct calculation to classify all possible forms of linear observed systems including process ODEs and algebraic observations on such multi-frame group through its automorphism structure, in comparison to the existing classification on two-frame groups relying on ingenious construction. Depth-camera inertial odometry with online extrinsics calibration is provided as an application.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is the classification problem of linear observation systems on multi - frame groups (MFG). Specifically, the author aims to: 1. **Construct multi - frame group structures**: Use the two - frame group (TFG) as a building block to construct multi - frame group structures through semi - direct products, covering all natural extensions. This solves the problem of the general group structure when simultaneously estimating states in multiple coupled coordinate systems. 2. **Systematically classify linear observation systems**: Utilize the automorphism group of MFG to propose a systematic method to directly calculate all possible forms of linear observation systems, including process ODEs and algebraic observation equations. Unlike methods that rely on clever constructions, this method provides broader applicability and theoretical guarantees. 3. **Provide practical application cases**: Take depth - camera inertial odometry as an example to show how to calibrate external parameters online and prove the effectiveness of the proposed multi - frame group - invariant extended Kalman filter (MFG - IEKF). Through these efforts, the paper fills the modeling gap from single - vehicle navigation to sensor systems involving multiple rigid - body motions and provides researchers and engineers with a powerful tool for designing robust and consistent observers. ### Formula Display For example, in the definition of multi - frame groups, matrix embedding and transformation chain structures are involved: \[ T=\begin{bmatrix} R&r\\ 0&I \end{bmatrix}\in GL(d + n + m,\mathbb{R}),\quad r = [x\ Ry] \] where \(R\in SO(d)\), \(r\in\mathbb{R}^{d\times(n + m)}\), \(x\in\mathbb{R}^{d\times n}\), \(y\in\mathbb{R}^{d\times m}\). Also, in the classification process, specific ODE forms are involved: \[ \dot{T}_0 = lS_0T_0+T_0rW_0 \] \[ \frac{d}{dt}[lT_j\cdots T_0]=lS_jlT_j\cdots T_0 + lT_j\cdots T_0lW_j \] \[ \frac{d}{dt}[T_0\cdots rT_j]=rS_jT_0\cdots rT_j+T_0\cdots rT_jrW_j \] These formulas ensure the complete classification and strict derivation of linear observation systems.

Classification of Linear Observed Systems on Multi-Frame Groups via Automorphisms

Developing Rigid Constraint for the Estimation of Pose and Structure from a Single Image

Direct Alignment with Generalized Correspondences: A Unified Framework for Structure-Based Visual Pose Estimation.

Robust and Efficient Vehicles Motion Estimation with Low-Cost Multi-Camera and Odometer-Gyroscope

Realization of odometry system using monocular vision

The Geometry of Navigation Problems

An almost globally convergent observer for visual SLAM without persistent excitation.

Local Optimized and Scalable Frame-to-model SLAM

On the Existence of Linear Observed Systems on Manifolds with Connection

Equivariant Filter Design for Kinematic Systems on Lie Groups

Multibody Grouping by Inference of Multiple Subspaces from High-Dimensional Data Using Oriented-Frames

Structure and Motion for Intelligent Vehicles Using an Uncalibrated Two-Camera System

A New Paradigm for Dealing With Manifold Structures in Visual Inertial Odometry by Using Stable Embedding

MCOV-SLAM: A Multicamera Omnidirectional Visual SLAM System

Feature-based Recursive Observer Design for Homography Estimation

A General Optimization-based Framework for Local Odometry Estimation with Multiple Sensors

Equivariant Filter for Tightly Coupled LiDAR-Inertial Odometry

Machine learning based state observer for discrete time systems evolving on Lie groups

Identification of a Moving Object's Velocity and Range with a Static-Moving Camera System

Guaranteed Performance Nonlinear Observer for Simultaneous Localization and Mapping

Nonlinear Stochastic Estimators on the Special Euclidean Group SE(3) using Uncertain IMU and Vision Measurements