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

The multiscale hermite transform for local orientation analysis.

José L Silvan-Cárdenas1, Boris Escalante-Ramírez

  • 1Centro de Investigación en Geogrfía y Geomática Ing. J. L. Tamayo (CentroGeo), Tlalpan, México, DF. jlsilvan@centrogeo.org.mx

IEEE Transactions on Image Processing : a Publication of the IEEE Signal Processing Society
|May 5, 2006
PubMed
Summary

This study introduces a novel multiscale image representation model inspired by the human visual system. It efficiently captures local image structures across various resolutions for advanced computer vision applications.

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

Automatic Segmentation of the Left Ventricle Through the Cardiac Cycle in Pediatric Echocardiography Videos Using SegFormer Architecture.

Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Annual International Conference·2025
Same author

Feature-based 3D+t descriptors of hyperactivated human sperm beat patterns.

Heliyon·2024
Same author

Robust cardiac segmentation corrected with heuristics.

PloS one·2023
Same author

3D+t feature-based descriptor for unsupervised flagellar human sperm beat classification.

Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Annual International Conference·2022
Same author

A multiphase texture-based model of active contours assisted by a convolutional neural network for automatic CT and MRI heart ventricle segmentation.

Computer methods and programs in biomedicine·2021
Same author

What is new in computer vision and artificial intelligence in medical image analysis applications.

Quantitative imaging in medicine and surgery·2021

Area of Science:

  • Computer Vision
  • Image Processing
  • Computational Neuroscience

Background:

  • Representing local differential structure at multiple resolutions is crucial for adaptive image processing.
  • Existing methods may not fully capture the nuances of natural image structures across scales.
  • The human visual system provides a biological model for efficient image representation.

Purpose of the Study:

  • To derive a multiscale model for natural image representation inspired by the human visual system.
  • To develop efficient operators for analyzing and synthesizing image structures at various resolutions.
  • To extend the model from one to two and three dimensions.

Main Methods:

  • Derivation of a one-dimensional multiscale model, extended to higher dimensions.

Related Experiment Videos

  • Utilizing Gaussian smoothing kernel derivatives for analysis and synthesis operators.
  • Employing generalized binomial filters for efficient coordinate system rotation in 2D.
  • Defining a discrete counterpart approximating continuous coordinate normalization via subsampling.

    • Main Results:

    • A robust multiscale model for representing local differential image structure.
    • Efficient operators derived from Gaussian derivatives, suitable for analysis and synthesis.
    • Generalized binomial filters enable efficient rotation and orientation estimation.
    • A discrete model approximating continuous properties for practical implementation.

    Conclusions:

    • The proposed multiscale model offers an efficient and biologically inspired approach to image representation.
    • The derived operators and methods facilitate accurate analysis of local image structures at multiple resolutions.
    • This work contributes to advancements in adaptive image processing and computer vision.