Fried M (2003)
Publication Language: English
Publication Type: Other publication type
Publication year: 2003
Publisher: Preprintserie des Mathematischen Instituts, Universität Freiburg
City/Town: Freiburg
Journal Issue: [03-26]
We present an adaptive finite element algorithm for the minimal
partition problem based on a level set formulation. Starting from the
piecewise Mumford-Shah model for image segmentation, we follow a
method proposed by Chan and Vese, which is able to detect exterior and
interior boundaries as well as objects, whose boundaries are not
necessarily defined by gradients. The algorithm stops automatically on
the object's boundary; quantities like volumes of the different
segments or of the interface between the segments could easily be
measured. Using adaptive methods, the computational effort could be
reduced in order to efficiently segment images in two and three
spatial dimensions. We present various numerical experiments for
synthetical and real images in two and three dimensions and compare
different adaptive approaches and initial interfaces.
APA:
Fried, M. (2003). Image segmentation using adaptive finite elements. Freiburg: Preprintserie des Mathematischen Instituts, Universität Freiburg.
MLA:
Fried, Michael. Image segmentation using adaptive finite elements. Freiburg: Preprintserie des Mathematischen Instituts, Universität Freiburg, 2003.
BibTeX: Download