Communication Dans Un Congrès Année : 2012

Spherical coordinates framed RGB color space dichromatic reflection model based image segmentation : application to wildland fire's outlines extraction

Résumé

Wildland fires represent a major risk for many countries over the world. For efficient fire fighting, the modeling and prediction of fire front propagation is a curial need. However, wildland fires' involves complex dynamics and mathematical modelling of such complex systems needs reliable information extraction from real situations, which is far from being a trivial task. Artificial Vision and Image Processing offer appealing potential toward reliable extraction of required information. In this paper we focus on flames' and fires' segmentation, dealing with the above-stated already open problem. The segmentation approach that we propose is based on dichromatic reflection model reformulated on a spherical interpretation of the RGB color space

Fichier non déposé

Dates et versions

hal-00761655 , version 1 (05-12-2012)

Identifiants

  • HAL Id : hal-00761655 , version 1

Citer

V. Amarger, Dominik M. Ramik, Christophe Sabourin, Kurosh Madani, Ramon Moreno, et al.. Spherical coordinates framed RGB color space dichromatic reflection model based image segmentation : application to wildland fire's outlines extraction. Third International Conference on Image Processing Theory, Tools and Applications, Oct 2012, Istanbul, Turkey. pp.19-24. ⟨hal-00761655⟩
155 Consultations
0 Téléchargements

Partager

  • More