Building on recent progress in modeling filter response statistics of natural mages we integrate a statistical model into a variational framework for image segmentation. Incorporated in asound probabilistic distance measure the model drives level sets toward meaningful segment at ions of complex textures and natural scenes. Despite its enhanced descriptive power our approach preserves the efficiency of level set based segmentation since each connected region comprises two model parameters only. We validate the statistical basis of our model on thousands of natural images and demonstrate that our approach outperforms recent variational segment at ion methods based on second-order statistics.
Additional information:
Das Dokument wird vom Publikationsserver der Universitätsbibliothek Mannheim bereitgestellt.
![[img]](https://madoc.bib.uni-mannheim.de/style/images/fileicons/application_pdf.png)
Search Authors in