Multi-Objective Optimization of Loop Closure Detection Parameters for Indoor 2D Simultaneous Localization and Mapping.

Dongxiao Han,Yuwen Li,Tao Song,Zhenyang Liu

doi:10.3390/s20071906

Abstract

Aiming at addressing the issues related to the tuning of loop closure detection parameters for indoor 2D graph-based simultaneous localization and mapping (SLAM), this article proposes a multi-objective optimization method for these parameters. The proposed method unifies the Karto SLAM algorithm, an efficient evaluation approach for map quality with three quantitative metrics, and a multi-objective optimization algorithm. More particularly, the evaluation metrics, i.e., the proportion of occupied grids, the number of corners and the amount of enclosed areas, can reflect the errors such as overlaps, blurring and misalignment when mapping nested loops, even in the absence of ground truth. The proposed method has been implemented and validated by testing on four datasets and two real-world environments. For all these tests, the map quality can be improved using the proposed method. Only loop closure detection parameters have been considered in this article, but the proposed evaluation metrics and optimization method have potential applications in the automatic tuning of other SLAM parameters to improve the map quality.

Highlights

As a key technology for autonomous navigation of mobile robots, simultaneous localization and mapping (SLAM) focuses on the problem of acquiring a spatial map of an environment while simultaneously localizing the robot using this map [1]
We evaluate the map quantitatively using the three metrics presented in multiple objectives for the optimization of the search area r and the amount of nodes g for Karto SLAM
The parameters involved in the proposed parameters in Karto SLAM algorithm according to the instructions of these datasets

Summary

Introduction

As a key technology for autonomous navigation of mobile robots, simultaneous localization and mapping (SLAM) focuses on the problem of acquiring a spatial map of an environment while simultaneously localizing the robot using this map [1]. Filter-based SLAM is mainly developed from the principle of recursive Bayesian estimation and is a problem of incremental, real-time data processing and robot pose correction. Due to its computational complexity and linearization treatment, the application of EKF SLAM has been limited by its scaling limitation and mapping inconsistence. To overcome some of these issues, particle filters have been proposed for SLAM by sampling from robot pose data associations, but the number of particles can grow large and the estimate can become inconsistent when mapping nested loops especially in large-scale environments [3,4,5]. Graph-based SLAM models the map as a sparse graph with constraints corresponding to the relation between robot motion and environment measurement

Methods

Results

Conclusion