Logo des Repositoriums
 

Real-time implementation of a random finite set particle filter

dc.contributor.authorReuter, Stephan
dc.contributor.authorDietmayer, Klaus
dc.contributor.authorHandrich, Sebastian
dc.contributor.editorHeiß, Hans-Ulrich
dc.contributor.editorPepper, Peter
dc.contributor.editorSchlingloff, Holger
dc.contributor.editorSchneider, Jörg
dc.date.accessioned2018-11-27T10:00:09Z
dc.date.available2018-11-27T10:00:09Z
dc.date.issued2011
dc.description.abstractIn scenarios characterized by a high object density, data association is a demanding task due to several ambiguities. Especially the assumption that all objects move independent from each other may lead to physically impossible predicted states, in which two objects are closer to each other than feasible. Thus, avoiding such impossible states may simplify the data association. Within the random finite set statistics it is possible to easily incorporate constraints concerning object states and to integrate them into a multi-target Bayes filter. A drawback of the random finite set statistics is its computational complexity, especially in the corrector step. In this contribution, a fast approximation for the calculation of the multi-target likelihood function is proposed. This approximation is used to implement a real-time random finite set particle filter on a graphical processing unit using real world sensor data.en
dc.identifier.isbn978-88579-286-4
dc.identifier.pissn1617-5468
dc.identifier.urihttps://dl.gi.de/handle/20.500.12116/18807
dc.language.isoen
dc.publisherGesellschaft für Informatik e.V.
dc.relation.ispartofINFORMATIK 2011 – Informatik schafft Communities
dc.relation.ispartofseriesLecture Notes in Informatics (LNI) - Proceedings, Volume P-192
dc.titleReal-time implementation of a random finite set particle filteren
dc.typeText/Conference Paper
gi.citation.endPage493
gi.citation.publisherPlaceBonn
gi.citation.startPage493
gi.conference.date4.-7. Oktober 2011
gi.conference.locationBerlin
gi.conference.sessiontitleRegular Research Papers

Dateien

Originalbündel
1 - 1 von 1
Lade...
Vorschaubild
Name:
493.pdf
Größe:
17.66 KB
Format:
Adobe Portable Document Format