Unified Representation of Geometric Primitives for Graph-SLAM Optimization Using Decomposed Quadrics

TL;DR

This paper introduces a unified quadric-based representation for geometric primitives in graph-SLAM, enhancing map compactness and robustness through a decomposed model that captures primitive properties more effectively.

Contribution

It proposes a novel decomposed quadric model for geometric primitives, improving efficiency and robustness in SLAM map optimization.

Findings

Decomposed quadrics outperform baseline in efficiency.

Decomposed formulation shows increased robustness to noise.

Framework produces compact, regularized maps in real-world tests.

Abstract

In Simultaneous Localization And Mapping (SLAM) problems, high-level landmarks have the potential to build compact and informative maps compared to traditional point-based landmarks. In this work, we focus on the parameterization of frequently used geometric primitives including points, lines, planes, ellipsoids, cylinders, and cones. We first present a unified representation based on quadrics, leading to a consistent and concise formulation. Then we further study a decomposed model of quadrics that discloses the symmetric and degenerated properties of a primitive. Based on the decomposition, we develop geometrically meaningful quadrics factors in the settings of a graph-SLAM problem. Then in simulation experiments, it is shown that the decomposed formulation has better efficiency and robustness to observation noises than baseline parameterizations. Finally, in real-world experiments, the proposed back-end framework is demonstrated to be capable of building compact and regularized maps.