Robust Online Trajectory Clustering without Computing Trajectory Distances

Michael Ulm (Vortragende:r), Norbert Brändle

Publikation: Beitrag in Buch oder TagungsbandBeitrag in Tagungsband mit PosterpräsentationBegutachtung

Abstract

We propose a novel trajectory clustering algorithm which is suitable for online processing of pedestrian or vehicle trajectories computed with a vision-based tracker. Our approach does not require defining distances between trajectories, and can thus process broken trajectories which are inevitable in most cases when object trackers are applied to real world video footage. Clusters are defined as smooth vector fields on a bounded connected set, and cluster distance is based on pairwise distances between vector sets. The results are illustrated on a trajectory set from the Edinburgh Informatics Forum Pedestrian Dataset, on a trajectory set from a public transport junction, and trajectories from an experimental setup in a corridor.
OriginalspracheEnglisch
TitelICPR 2012
Seiten2270-2273
Seitenumfang4
PublikationsstatusVeröffentlicht - 2012
VeranstaltungInternational Conference on Pattern Recognition 2012 -
Dauer: 11 Nov. 201215 Nov. 2012

Konferenz

KonferenzInternational Conference on Pattern Recognition 2012
Zeitraum11/11/1215/11/12

Research Field

  • Ehemaliges Research Field - Mobility Systems

Fingerprint

Untersuchen Sie die Forschungsthemen von „Robust Online Trajectory Clustering without Computing Trajectory Distances“. Zusammen bilden sie einen einzigartigen Fingerprint.

Diese Publikation zitieren