Probabilistic Localization of a Mobile Robot Based on the Sensor Fusion of a Laser Scanner and a Monocular Camera

EasyChair Preprint 10971

Abstract

In the context of indoor localization there is a massive use of LiDAR-based approaches due to their real-time performance and high accuracy, and it was perceived that these methods present difficulties in symmetrical environments and environments with lack of longitudinal reference. This paper deals with the proposition of a localization problem approach for a mobile robot in indoor environments using scanning and image sensory. Considering the existence of a map of the environment containing fiducial markers, it brings the image sensor to overcome LiDAR exteroceptive perception limitations. The results obtained indicated gains of up to 19.35% in the accuracy of determining the location of the system in relation to LiDAR-based methods in scenarios with low range LiDAR.

Keyphrases: Autonomous Vehicle Navigation, Fidutial Marks, indoor localization, particle filter, sensor fusion

@booklet{EasyChair:10971,
  author    = {Lucas Costa and Larissa Da Silva and Antonio Lima},
  title     = {Probabilistic Localization of a Mobile Robot Based on the Sensor Fusion of a Laser Scanner and a Monocular Camera},
  howpublished = {EasyChair Preprint 10971},
  year      = {EasyChair, 2023}}