Accurate and Robust Teach and Repeat Navigation by Visual Place Recognition: A CNN Approach

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F68407700%3A21730%2F20%3A00348826" target="_blank" >RIV/68407700:21730/20:00348826 - isvavai.cz</a>

Výsledek na webu

<a href="https://doi.org/10.1109/IROS45743.2020.9341764" target="_blank" >https://doi.org/10.1109/IROS45743.2020.9341764</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1109/IROS45743.2020.9341764" target="_blank" >10.1109/IROS45743.2020.9341764</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Accurate and Robust Teach and Repeat Navigation by Visual Place Recognition: A CNN Approach

Popis výsledku v původním jazyce

We propose a visual teach-and-repeat navigation system, SSM-Nav, which is based on the output of the re cently introduced SSM visual place recognition methodology. During the teach phase, a teleoperated wheeled robot stores in a database features of images taken along an arbitrary route. During the repeat phase or navigation, a CNN-based comparison of each captured image is performed against the database. With the help of a particle filter, the best location’s image is selected and its horizontal offset with respect to the current scene used to correct the steering of the robot and to navigate. Indoor tests in our lab show a maximum error of less than 10 cm and excellent robustness to perturbations such as drastic changes in illumination, lateral displacements, different starting positions or even kidnapping. Preliminary outdoor tests on a 0.22 Km route show promising results, with an estimated maximum error of less than 25 cm.

Název v anglickém jazyce

Accurate and Robust Teach and Repeat Navigation by Visual Place Recognition: A CNN Approach

Popis výsledku anglicky

We propose a visual teach-and-repeat navigation system, SSM-Nav, which is based on the output of the re cently introduced SSM visual place recognition methodology. During the teach phase, a teleoperated wheeled robot stores in a database features of images taken along an arbitrary route. During the repeat phase or navigation, a CNN-based comparison of each captured image is performed against the database. With the help of a particle filter, the best location’s image is selected and its horizontal offset with respect to the current scene used to correct the steering of the robot and to navigate. Indoor tests in our lab show a maximum error of less than 10 cm and excellent robustness to perturbations such as drastic changes in illumination, lateral displacements, different starting positions or even kidnapping. Preliminary outdoor tests on a 0.22 Km route show promising results, with an estimated maximum error of less than 25 cm.

Druh

D - Stať ve sborníku

CEP obor

—

OECD FORD obor

20204 - Robotics and automatic control

Projekt

Výsledek vznikl pri realizaci vícero projektů. Více informací v záložce Projekty.

Návaznosti

P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)

Rok uplatnění

2020

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název statě ve sborníku

Proceedings of the 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems

ISBN

978-1-7281-6212-6

ISSN

2153-0858

e-ISSN

—

Počet stran výsledku

7

Strana od-do

6018-6024

Název nakladatele

IEEE Computer Society

Místo vydání

Los Alamitos

Místo konání akce

Las Vegas

Datum konání akce

24. 10. 2020

Typ akce podle státní příslušnosti

WRD - Celosvětová akce

Kód UT WoS článku

—