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 model for size- and rotation-invariant pattern processing in the visual system.

H J Reitboeck, J Altmann

    Biological Cybernetics
    |January 1, 1984
    PubMed
    Summary

    This study presents a neural network model for visual pattern recognition. It uses a log-polar transform and R-transform to achieve size and rotation invariance, mimicking retinal processing for object recognition.

    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

    Fertility-sparing strategy in a rare case of uterine adenosarcoma and consecutive spontaneous pregnancy and livebirth.

    Archives of gynecology and obstetrics·2026
    Same author

    Severe hemorrhage in choriocarcinoma: three scenarios and treatment strategies.

    Gynecologic oncology reports·2025
    Same author

    Fertility-sparing strategy in a rare case of highly malignant Dicer-1-associated sarcoma of the cervix.

    Archives of gynecology and obstetrics·2024
    Same author

    A dramatic response to checkpoint inhibitor in a woman with small cell carcinoma of the hypercalcemic type of the ovary.

    Gynecologic oncology·2023
    Same author

    Lifting the veil of secrecy: maternal and neonatal outcome of oocyte donation pregnancies in Germany.

    Archives of gynecology and obstetrics·2021
    Same author

    Body mass and growth rates in a wild primate population.

    Oecologia·2017

    Area of Science:

    • Computational Neuroscience
    • Visual System Modeling

    Background:

    • Retinal space mapping to the striate cortex approximates a log-polar function in mammals.
    • This mapping is hypothesized to be crucial for scale- and rotation-invariant pattern recognition.

    Purpose of the Study:

    • To propose and evaluate a computational model for object recognition independent of size, orientation, and position.
    • To implement translation invariance using the R-transform, avoiding complex Fourier transform computations.

    Main Methods:

    • Utilized a log-polar transform to convert scaling and rotation into translations.
    • Employed "Mexican hat" filters with linearly increasing diameter for logarithmic space distortion and differentiation.
    • Implemented translation invariance via the R-transform, suitable for simple neural networks.

    Main Results:

    • The model demonstrates object recognition independent of size, orientation, and position, with some exceptions.
    • The R-transform offers a computationally simpler alternative to Fourier transforms for achieving size invariance.

    Conclusions:

    • The proposed model provides a viable mechanism for scale- and rotation-invariant pattern recognition in neural systems.
    • The R-transform is an efficient method for achieving translation invariance in computational models of vision.

    Related Experiment Videos