Eventos Anais de eventos
CONEM 2018
X Congresso Nacional de Engenharia Mecânica
A Pose-Graph Optimization tool for MATLAB
Submission Author:
João Carlos Virgolino Soares , RJ
Co-Authors:
João Carlos Virgolino Soares, Marco Antonio Meggiolaro
Presenter: João Carlos Virgolino Soares
doi://10.26678/ABCM.CONEM2018.CON18-0341
Abstract
Several problems in mobile robotics need probabilistic formulations to handle the inherent uncertainty of motion and sensor measurements. Graph-SLAM is a probabilistic approach to the simultaneous localization and mapping problem that is based on maximum likelihood estimation and non-linear least squares optimization. It consists in generating a graph from the poses of the robot and from the constraints of measurements between poses, followed by the optimization of this graph to obtain a consistent trajectory. This paper presents the development and implementation of a pose-graph optimization tool for MATLAB. A pose quaternion representation is used and a manifold optimization approach is applied to handle attitude representation problems in the optimization. The system is evaluated using benchmark datasets available in the literature.
Keywords
Graph-SLAM, Non-linear least squares, Probabilistic Robotics
