Faculty of Informatics Vienna University of Technology Institute of Computer Aided Automation PRIP Home PRIP Home
Personal tools
You are here: Home Teaching Sommersemester (SS) DiplDiss Presentations Interactive Image-Segmentation on a MST-based Graph-Pyramid (M. Gerstmayer)

Interactive Image-Segmentation on a MST-based Graph-Pyramid (M. Gerstmayer)

What
  • Master Thesis Presentation
When May 18, 2010
from 04:10 pm to 04:20 pm
Where Seminarraum 183/2
Add event to calendar vCal
iCal
Computational Imaging and Vison, in special Image Segmentation, has various applications and depending on the domain in consideration, just as many concepts or approaches that can in uence the algorithm. Inspired from medical applications and the MST-framework, we integrate here the human user and therefore combine external knowledge with an advanced representation. The former guides the algorithm to a result by placing modifying-actions, the latter mentioned framework is a 'Hierarchical graph-based Image Segmentation Method' called MST Pyramid (developed by Adrian Ion, Yll Haxhimusa and Walter G. Kropatsch 1). The internal representation, calculation and interaction-processing (merge/protect faces), is based on combinatorial maps, which encode the structure efficiently. Output of the developed system is a user-modi ed (hierarchical) segmentation where for example you can cut out (or decompose) a desired object. Besides is not limited to this area: Image Labelling or Annotation is one intended future use, machine-learning another. The thesis will cover theory and problems concerning the interplay of both parts and a comparison to alternative methods used in practice, hoping to make a valuable contribution to this area of resarch.
Document Actions