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 Video

Updated: Jul 7, 2026

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
04:48

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Published on: November 30, 2022

Cluster validation for unsupervised stochastic model-based image segmentation.

D A Langan1, J W Modestino, J Zhang

  • 1Center for Image Processing Research, Electrical, Computer, and Systems Engineering Department, Rensselaer Polytechnic Institute, Troy, NY 12180, USA.

IEEE Transactions on Image Processing : a Publication of the IEEE Signal Processing Society
|February 13, 2008
PubMed
Summary

Determining the number of classes in unsupervised image segmentation is challenging. A new model-fitting technique robustly identifies the correct number of classes, outperforming traditional information-theoretic criteria.

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

Texture discrimination based upon an assumed stochastic texture model.

IEEE transactions on pattern analysis and machine intelligence·2011
Same author

A maximum likelihood approach to texture classification.

IEEE transactions on pattern analysis and machine intelligence·2011
Same author

Operational rate-distortion performance for joint source and channel coding of images.

IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2008
Same author

Reliable transmission of high-quality video over ATM networks.

IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2008
Same author

In vivo clonotypic regulation of human myelin basic protein-reactive T cells by T cell vaccination.

Journal of immunology (Baltimore, Md. : 1950)·1995
Same author

Superantigen reactivity of gamma delta T cell clones isolated from patients with multiple sclerosis and controls.

Cellular immunology·1995

Area of Science:

  • Computer Vision
  • Image Analysis
  • Machine Learning

Background:

  • Unsupervised image segmentation is crucial for image analysis when training data is unavailable.
  • A key challenge is determining the correct number of classes within an image, known as the cluster validation problem.

Purpose of the Study:

  • To investigate the cluster validation problem in unsupervised image segmentation using the expectation-maximization (EM) algorithm.
  • To evaluate the performance of information-theoretic criteria (ICs) for cluster validation with EM-based segmentation.
  • To propose and validate a novel model-fitting technique for cluster validation.

Main Methods:

  • Applied expectation-maximization (EM) algorithm for unsupervised image segmentation.
  • Evaluated established information-theoretic criteria (ICs) for determining the number of classes.

More Related Videos

From Voxels to Knowledge: A Practical Guide to the Segmentation of Complex Electron Microscopy 3D-Data
12:08

From Voxels to Knowledge: A Practical Guide to the Segmentation of Complex Electron Microscopy 3D-Data

Published on: August 13, 2014

Related Experiment Videos

Last Updated: Jul 7, 2026

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
04:48

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Published on: November 30, 2022

From Voxels to Knowledge: A Practical Guide to the Segmentation of Complex Electron Microscopy 3D-Data
12:08

From Voxels to Knowledge: A Practical Guide to the Segmentation of Complex Electron Microscopy 3D-Data

Published on: August 13, 2014

  • Developed a new model-fitting technique by modeling the log-likelihood functional as an exponential function of the number of classes.

    • Main Results:

    • Information-theoretic criteria (ICs) often yield inappropriate solutions due to penalty term domination in EM-based segmentation.
    • The proposed model-fitting technique demonstrated robustness in determining the number of classes.
    • Experimental results showed the new technique outperforms traditional ICs on both synthetic and real-world imagery.

    Conclusions:

    • The proposed model-fitting technique offers a superior solution to the cluster validation problem in unsupervised image segmentation compared to existing ICs.
    • This method provides a more reliable way to estimate the number of classes, enhancing the accuracy of segmentation algorithms.