Temporal Constraints for Video Interpretation

LES AUTEURS Van-Thinh VU, François Bremond, Monique Thonnat

RESUME: This paper presents an original approach for temporal scenario recog-nition for video interpretation. We propose a declarative model to represent sce-narios and we use a logic-based approach to recognize pre-defined scenario models. To speed up the recognition process, we propose a new method to pro-cess the temporal operators of interval algebra and a method to extend the time interval of recognized scenarios. We have tested our representation formalism and the inference engine on two real video sequences.

Mots clé: video interpretation, on-line human behaviour recognition, cognitive vision, human behaviour representation

Pour télécharger cet article,  cliquez ici

BibTeX reference:

@INPROCEEDINGS{ECAI2002,
   AUTHOR    = {Van-Thinh VU, François BREMOND and Monique THONNAT},
   ADDRESS   = {Lyon, FRANCE},
   BOOKTITLE = {15th European Conference on Artificial Intelligence},
   TITLE     = {Temporal Constraints for Video Interpretation},