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 Markov model for blind image separation by a mean-field EM algorithm.

Anna Tonazzini1, Luigi Bedini, Emanuele Salerno

  • 1Istituto di Scienza e Tecnologie dell'Informazione, I-56124 Pisa, Italy. anna.tonazzini@isti.cnr.it

IEEE Transactions on Image Processing : a Publication of the IEEE Signal Processing Society
|February 17, 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

Complementing Hi-C information for 3D chromatin reconstruction by ChromStruct.

Frontiers in bioinformatics·2024
Same author

Restoration and content analysis of ancient manuscripts via color space based segmentation.

PloS one·2023
Same author

AI applications in functional genomics.

Computational and structural biotechnology journal·2021
Same author

Analysis of Diagnostic Images of Artworks and Feature Extraction: Design of a Methodology.

Journal of imaging·2021
Same author

Integration of Multiple Resolution Data in 3D Chromatin Reconstruction Using <i>ChromStruct</i>.

Biology·2021
Same author

Analytical and mathematical methods for revealing hidden details in ancient manuscripts and paintings: A review.

Journal of advanced research·2019
Same journal

Style-Aware Contrastive Test-Time Adaptation: A Dual-Cache Model for Robust Vision-Language Alignment.

IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2026
Same journal

Semantic Frame Interpolation.

IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2026
Same journal

Physics-Guided Cross-Modal Decoupling with Test-Time Adaptation for Hyperspectral Image Restoration.

IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2026
Same journal

Change-Prior-Guided Unsupervised Change Detection of Heterogeneous Remote Sensing Images.

IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2026
Same journal

AgonicDreamer: Enhancing Multi-View Consistency in Text-to-3D Generation via Rectified Score Distillation.

IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2026
Same journal

BiCM-Prompt: Bidirectional Cross-Modal Prompt Tuning for Class-Incremental Learning on Multisource Remote Sensing Images.

IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2026
See all related articles

This study introduces a Bayesian approach for blind image separation from noisy mixtures using Markov Random Field (MRF) models. The method enhances robustness against noise and allows separation of cross-correlated sources.

Area of Science:

  • Computer Vision
  • Statistical Signal Processing
  • Machine Learning

Background:

  • Blind source separation (BSS) aims to recover original signals from mixtures without prior knowledge.
  • Traditional BSS methods often assume statistical independence of sources, limiting their applicability.
  • Image reconstruction tasks benefit from models capturing local image correlations.

Purpose of the Study:

  • To develop a robust Bayesian framework for blind separation of noisy images.
  • To incorporate local image correlations using Markov Random Field (MRF) models.
  • To enable simultaneous separation and edge detection in images.

Main Methods:

  • Formulated blind image separation as a Bayesian estimation problem.
  • Employed Markov Random Field (MRF) image models to capture local correlations and preserve discontinuities.

Related Experiment Videos

  • Utilized an expectation-maximization (EM) algorithm with mean field approximation for parameter and source estimation.

    • Main Results:

    • The proposed MRF-based Bayesian approach demonstrated increased robustness against noise, including space-variant noise.
    • Simultaneous separation and edge detection were achieved.
    • The method successfully separated cross-correlated sources when the model accurately reflected source characteristics.

    Conclusions:

    • Bayesian blind image separation using MRF models offers a flexible and robust alternative to traditional methods.
    • Accounting for local image autocorrelation improves performance in noisy conditions.
    • The framework relaxes the strict independence assumption, broadening the scope of separable sources.