A context representation for surveillance systems

Authors: F. Bremond and M. Thonnat,

Abstract:

In this paper we address the issue of context representation for scene interpretation systems. Although contextual information is often essential for these systems, its effective utilization remains elusive. After proposing a definition of context, the first part of this paper explainsthe reasons why scene interpretation systems especially need context. Then, wedescribe a global context representation, expressed under a common formalism. This paper ends with an example of effective utilization of the proposed context representation.

Keywords: Scene interpretation, Contextual information, Context representation.

To get this article, click here

BibTeX reference:

@INPROCEEDINGS{ECCV96,
  AUTHOR    = {Bremond , F and Thonnat , M},
  BOOKTITLE = {ECCV Worshop on Conceptual Descriptions from Images},
  MONTH     = {April},
  TITLE     = {A context representation for surveillance systems},
  YEAR      = {1996},
  LOCATION  = {Cambridge}
}

Dernière mise à jour: 15/03/01
Agnes.Cortell@sophia.inria.fr