Crowd Flow Characterization with Optimal Control Theory

Abstract : Analyzing the crowd dynamics from video sequences is an open challenge in computer vision. Under a high crowd density assumption, we characterize the dynamics of the crowd flow by two related information: velocity and a disturbance potential which accounts for several elements likely to disturb the flow (the density of pedestrians, their interactions with the flow and the environment). The aim of this paper to simultaneously estimate from a sequence of crowded images those two quantities. While the velocity of the flow can be observed directly from the images with traditional techniques, this disturbance potential is far more trickier to estimate. We propose here to couple, through optimal control theory, a dynamical crowd evolution model with observations from the image sequence in order to estimate at the same time those two quantities from a video sequence. For this purpose, we derive a new and original continuum formulation of the crowd dynamics which appears to be well adapted to dense crowd video sequences. We demonstrate the efficiency of our approach on both synthetic and real crowd videos.
Type de document :
Communication dans un congrès
Ninth Asian Conference on Computer Vision (ACCV 2009), Jul 2009, Xi'an, China. pp.279--290, 2009
Liste complète des métadonnées


https://hal.archives-ouvertes.fr/hal-00493443
Contributeur : Nicolas Courty <>
Soumis le : mardi 22 juin 2010 - 10:04:00
Dernière modification le : mardi 13 octobre 2015 - 01:09:14
Document(s) archivé(s) le : vendredi 24 septembre 2010 - 17:37:18

Fichier

Crowd.pdf
Fichiers produits par l'(les) auteur(s)

Identifiants

  • HAL Id : hal-00493443, version 1

Collections

Citation

Pierre Allain, Nicolas Courty, Thomas Corpetti. Crowd Flow Characterization with Optimal Control Theory. Ninth Asian Conference on Computer Vision (ACCV 2009), Jul 2009, Xi'an, China. pp.279--290, 2009. <hal-00493443>

Partager

Métriques

Consultations de
la notice

366

Téléchargements du document

379