Fusão de Dados para a Estimativa de Localização de Robôs Móveis no Sistema MobiLysa

Matheus D. Oliveira; Vinícius H. Izabel; Leonardo A. Silva; Alexandre P. Do Carmo; Flávio Machado; Raquel F. Vassallo

doi:10.20906/SBAI-SBSE-2023/3820

Matheus D. Oliveira Departamento de Engenharia Elétrica, Universidade Federal do Espírito Santo - ES
Vinícius H. Izabel Departamento de Engenharia Elétrica, Universidade Federal do Espírito Santo - ES
Leonardo A. Silva Coord. de Eletrotécnica, Instituto Federal do Espírito Santo, ES
Alexandre P. Do Carmo Coord. de Eletrotécnica, Instituto Federal do Espírito Santo, ES
Flávio Machado IS Vision, Vitória, ES
Raquel F. Vassallo Departamento de Engenharia Elétrica, Universidade Federal do Espírito Santo - ES

DOI: https://doi.org/10.20906/SBAI-SBSE-2023/3820

Keywords: Data fusion, EKF, Computer vision, Localization, Intelligent Spaces

Abstract

The MobiLysa System aims to integrate the Lysa robot into an intelligent environment to assist visually impaired or low vision people to move in indoor spaces. This paper presents the robot’s localization module, which uses sensor fusion of two sources of position and orientation data: multicamera three-dimensional reconstruction by computer vision and mapbased estimation using the Monte Carlo algorithm. The data fusion is performed by the Extended Kalman Filter, increasing the accuracy and precision of the estimates. The paper describes in detail the data fusion and presents experiments to validate the proposed approach.