An active contour model without edges

Tony Chan, Luminita Vese

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

310 Scopus citations

Abstract

In this paper, we propose a new model for active contours to detect objects in a given image, based on techniques of curve evolution, Mumford-Shah functional for segmentation and level sets. Our model can detect objects whose boundaries are not necessarily defined by gradient. The model is a combination between more classical active contour models using mean curvature motion techniques, and the Mumford-Shah model for segmentation. We minimize an energy which can be seen as a particular case of the so-called minimal partition problem. In the level set formulation, the problem becomes a \mean-curvature flow"-like evolving the active contour, which will stop on the desired boundary. However, the stopping term does not depend on the gradient of the image, as in the classical active contour models, but is instead related to a particular segmentation of the image. Finally, we will present various experimental results and in particular some examples for which the classical snakes methods based on the gradient are not applicable.

Original languageEnglish (US)
Title of host publicationScale-Space Theories in Computer Vision - 2nd International Conference, Scale-Space 1999, Proceedings
EditorsMads Nielsen, Peter Johansen, Ole Fogh Olsen, Joachim Weickert
PublisherSpringer Verlag
Pages141-151
Number of pages11
ISBN (Print)354066498X, 9783540664987
StatePublished - Jan 1 1999
Event2nd International Conference on Scale-Space Theories in Computer Vision, 1999 - Corfu, Greece
Duration: Sep 26 1999Sep 27 1999

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume1682
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference2nd International Conference on Scale-Space Theories in Computer Vision, 1999
CountryGreece
CityCorfu
Period09/26/9909/27/99

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Fingerprint

Dive into the research topics of 'An active contour model without edges'. Together they form a unique fingerprint.

Cite this