Practical high-speed motion sensing: event cameras vs. global shutter

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F68407700%3A21730%2F20%3A00338030" target="_blank" >RIV/68407700:21730/20:00338030 - isvavai.cz</a>

Nalezeny alternativní kódy

RIV/60162694:G43__/20:00555949

Výsledek na webu

<a href="http://data.vicos.si/cvww20/26.pdf" target="_blank" >http://data.vicos.si/cvww20/26.pdf</a>

DOI - Digital Object Identifier

—

Jazyk výsledku

angličtina

Název v původním jazyce

Practical high-speed motion sensing: event cameras vs. global shutter

Popis výsledku v původním jazyce

We designed two maximally simplified and controlled experiments to test event-based and frame-based cameras in the price category affordable for an ordinary university research lab. First, we put common ArUco markers on a rotating disk and observed them by both types of cameras. We reconstructed the image from an event camera using a publicly available state-of-the-art algorithm and compared the ArUco marker recognition reliability. Surprisingly, our results suggest that the ability of the tested event camera to recognise quickly moving markers is inferior to an affordable $1000$~fps frame-based camera. In the second experiment, we let the cameras observe a freely flying subsonic pistol projectile. A very expensive $20000+$~fps camera provided ground-truth images, as the acquisition rate of the affordable frame-based camera was insufficient. Although event camera data was partially corrupt, it still allowed us to estimate the position of the small projectile every $10$~$mu$s, when the projectile translated mostly along the event camera pixel rows.

Název v anglickém jazyce

Practical high-speed motion sensing: event cameras vs. global shutter

Popis výsledku anglicky

We designed two maximally simplified and controlled experiments to test event-based and frame-based cameras in the price category affordable for an ordinary university research lab. First, we put common ArUco markers on a rotating disk and observed them by both types of cameras. We reconstructed the image from an event camera using a publicly available state-of-the-art algorithm and compared the ArUco marker recognition reliability. Surprisingly, our results suggest that the ability of the tested event camera to recognise quickly moving markers is inferior to an affordable $1000$~fps frame-based camera. In the second experiment, we let the cameras observe a freely flying subsonic pistol projectile. A very expensive $20000+$~fps camera provided ground-truth images, as the acquisition rate of the affordable frame-based camera was insufficient. Although event camera data was partially corrupt, it still allowed us to estimate the position of the small projectile every $10$~$mu$s, when the projectile translated mostly along the event camera pixel rows.

Druh

D - Stať ve sborníku

CEP obor

—

OECD FORD obor

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Projekt

<a href="/cs/project/EF15_003%2F0000470" target="_blank" >EF15_003/0000470: Robotika pro Průmysl 4.0</a><br>

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 25th Computer Vision Winter Workshop Conference February 3-5, 2020, Rogaška Slatina, Slovenia

ISBN

978-961-90901-9-0

ISSN

—

e-ISSN

—

Počet stran výsledku

9

Strana od-do

55-63

Název nakladatele

Slovenian Pattern Recognition Society

Místo vydání

Ljubljana

Místo konání akce

Rogaška Slatina

Datum konání akce

3. 2. 2020

Typ akce podle státní příslušnosti

WRD - Celosvětová akce

Kód UT WoS článku

—