Monocular Odometry for Underwater Vehicles with Online Estimation of the Scale Factor

Vincent Creuze 1
1 ICAR - Image & Interaction
LIRMM - Laboratoire d'Informatique de Robotique et de Microélectronique de Montpellier
Abstract : This paper introduces a new visual odometry method for underwater vehicles. It is based on images grabbed by a monocular video camera, and aided by inertial and pressure measurements. This approach offers several advantages. Firstly, it is compact and runs very fast, even with limited computational resources. This allows to embed it on very small vehicles. Secondly, and probably most importantly, the method is able to estimate online the scale factor of the observed scene, thanks to the combined measures of a low-cost IMU and a pressure sensor. The paper ends with an experimental validation onboard the Leonard underwater vehicle.
Complete list of metadatas

https://hal-lirmm.ccsd.cnrs.fr/lirmm-01567463
Contributor : Vincent Creuze <>
Submitted on : Sunday, July 23, 2017 - 11:22:24 PM
Last modification on : Thursday, May 24, 2018 - 3:59:23 PM

File

IFAC 2017 CREUZE version HAL.p...
Files produced by the author(s)

Identifiers

  • HAL Id : lirmm-01567463, version 1

Citation

Vincent Creuze. Monocular Odometry for Underwater Vehicles with Online Estimation of the Scale Factor. IFAC 2017 World Congress, Jul 2017, Toulouse, France. ⟨lirmm-01567463⟩

Share

Metrics

Record views

234

Files downloads

325