Filtrage conditionnel pour le suivi de points dans des sequences d'images

Elise Arnaud 1 Etienne Mémin 1 Bruno Cernuschi-Frias 2
1 VISTA - Vision spatio-temporelle et active
IRISA - Institut de Recherche en Informatique et Systèmes Aléatoires, Inria Rennes – Bretagne Atlantique
Abstract : The approach we investigate for point tracking combines within a stochastic filtering framework a dynamic model relying on the optical flow constraint and measurements provided by a matching technique. Based on this, two kinds of simple trackers are devised whose dynamic and measurements depend on image data. The first one is a linear tracker particularly well-suited to image sequences exhibiting global dominant motion situations. In this context, a conditional Kalman filter is derived through the use of a conditional linear minimum variance estimator. The second one is a nonlinear tracker built from a conditional particle filter. The two trackers allow us to deal with trajectories exhibiting abrupt changes. We present some experimental results on real-world image sequences and compare them to the Shi-Tomasi-Kanade feature tracker.
Document type :
Conference papers
Complete list of metadatas

Cited literature [21 references]  Display  Hide  Download


https://hal.inria.fr/inria-00590168
Contributor : Team Perception <>
Submitted on : Tuesday, May 3, 2011 - 9:29:41 AM
Last modification on : Tuesday, October 1, 2019 - 3:18:07 PM
Long-term archiving on: Thursday, August 4, 2011 - 3:02:03 AM

Files

Orasis03Arnaud.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : inria-00590168, version 1

Citation

Elise Arnaud, Etienne Mémin, Bruno Cernuschi-Frias. Filtrage conditionnel pour le suivi de points dans des sequences d'images. Journée francophones des jeunes chercheurs en vision par ordinateur (ORASIS '03), May 2003, Gérardmer, France. ⟨inria-00590168⟩

Share

Metrics

Record views

218

Files downloads

171