Dynamic Texture Segmentation

We address the problem of segmenting a sequence of images of natural scenes into disjoint regions that are characterized by constant spatio-temporal statistics. We model the spatio-temporal dynamics in each region by Gauss-Markov models, and infer the model parameters as well as the boundary of the regions in a variational optimization framework. Numerical results demonstrate that – in contrast to purely texture-based segmentation schemes – our method is effective in segmenting regions that differ in their dynamics even when spatial statistics are identical.

• Type: Conference
• Year: 2003
• Where: ICCV
• Authors: Gianfranco Doretto, Daniel Cremers, Paolo Favaro, Stefano Soatto
• Keywords: Computer Vision, Constant Spatio-temporal Statistics, Disjoint Regions, ICCV 2003, Variational Optimization Framework