Fluminance publications


Jump to : Abstract | Contact | BibTex reference | EndNote reference |


C. Avenel, E. Mémin, P. Pérez. Tracking levels representation driven by a stochastic dynamics. In 7th International Conference on Curves and Surfaces, Poster presentation, Avignon, France, June 2010.


The joint analysis of motions and deformations is crucial in a number of computer vision applications. In this paper, we introduce a non-linear stochastic ltering technique to track the state of a free curve. The approach we propose is implemented through a particle lter which includes color measurements character- izing the target and the background respectively. We design a continuous-time dynamics that allows us to infer inter-frame deformations. The curve is de ned by an implicit level-set representation and the stochastic dynamics is expressed on the level-set function. It takes the form of a stochastic di erential equation with Brownian motion of low dimension. Speci c noise models lead to traditional evolution laws based on mean curvature motions, while other forms lead to new evolution laws with di erent smoothing behaviors. In these evolution models, we propose to combine local motion information extracted from the images, some velocity induce by the curve displacement and an incertitude modeling of the dynamics.The associated lter we propose for curve tracking thus belongs to the family of conditional particle lters. Its capabilities are demonstrated on various sequences with highly deformable objects.


Etienne Mémin

BibTex Reference

   Author = {Avenel, C. and Mémin, E. and Pérez, P.},
   Title = {Tracking levels representation driven by a stochastic dynamics},
   BookTitle = {7th International Conference on Curves and Surfaces},
   Address = {Avignon, France},
   Month = {June},
   Year = {2010}

EndNote Reference [help]

Get EndNote Reference (.ref)

This page has been automatically generated by bib2html.

logo of irmar logo of cnrs logo of inrialogo of irstea

Copyright 2009 © Fluminance Team - INRIA - IRSTEA

Webmaster mailto