Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

A versatile segmentation procedure.

Walter Vanzella1, Vincent Torre

  • 1Department of Neurobiology, International School for Advanced Studies, Trieste 34014, Italy. vanzella@sissa.it

IEEE Transactions on Systems, Man, and Cybernetics. Part B, Cybernetics : a Publication of the IEEE Systems, Man, and Cybernetics Society
|April 11, 2006
PubMed
Summary
This summary is machine-generated.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Exosomal TNF-α mediates voltage-gated Na+ channel 1.6 overexpression and contributes to brain tumor-induced neuronal hyperexcitability.

The Journal of clinical investigation·2024
Same author

Mechanisms of Glioblastoma Replication: Ca2+ Flares and Cl- Currents.

Molecular cancer research : MCR·2024
Same author

Structural heterogeneity of the ion and lipid channel TMEM16F.

Nature communications·2024
Same author

Unfolding and identification of membrane proteins in situ.

eLife·2022
Same author

The dual action of glioma-derived exosomes on neuronal activity: synchronization and disruption of synchrony.

Cell death & disease·2022
Same author

2D MXene interfaces preserve the basal electrophysiology of targeted neural circuits.

Nanoscale·2022

A novel image segmentation method uses a self-adaptive Mumford-Shah functional for natural images. This approach automatically adjusts parameters for scale and contrast, enabling versatile segmentation for tasks like image recognition.

Area of Science:

  • Computer Vision
  • Image Processing
  • Computational Mathematics

Background:

  • Image segmentation is crucial for computer vision tasks.
  • Existing methods often struggle with varying local scales and contrast in natural images.
  • Parameter tuning in segmentation algorithms can be complex and time-consuming.

Purpose of the Study:

  • To propose a new, versatile method for natural image segmentation.
  • To develop an approach where key parameters adapt automatically to image characteristics.
  • To enable multi-resolution segmentation by adjusting a single parameter.

Main Methods:

  • Regularization of original images using a self-adaptive Mumford-Shah functional.
  • Automatic adaptation of smoothness (alpha) and fidelity (gamma) parameters to local image properties.

Related Experiment Videos

  • Generation of a piecewise constant image (sN) from the regularized image (u).
  • Merging of regions by minimizing a functional dependent on parameter 'n' for resolution control.

    • Main Results:

    • The method produces a piecewise constant image representing segmentation with N closed regions.
    • Preservation of thin bars and trihedral junctions in the segmented image.
    • Adjusting parameter 'n' allows for segmentations at different resolutions, from coarse (n=1) to fine (hundreds of regions).
    • The method was validated on datasets with ground truth segmentations.

    Conclusions:

    • The proposed image segmentation method is versatile and robust.
    • It relies on a single parameter 'n' for resolution control.
    • The technique is suitable for higher-level image processing tasks such as categorization, recognition, and scene understanding.