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

Planarity and segmentation in stereoscopic matching.

G Mitchison1

  • 1King's College Research Centre, King's College, University of Cambridge, UK.

Perception
|January 1, 1988
PubMed
Summary

Human vision solves stereo correspondence by segmenting images using coarse features. Planar approximations within segments help determine depth, especially in noisy images with periodic patterns.

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

Computing protein function.

Nature biotechnology·2000
Same author

Is there a phylogenetic signal in prokaryote proteins?

Journal of molecular evolution·1999
Same author

Making family trees from gene families.

Nature genetics·1999
Same author

Where is the mind's eye? Visual perception.

Current biology : CB·1996
Same author

REM sleep and neural nets.

Behavioural brain research·1995
Same author

Maximum discrimination hidden Markov models of sequence consensus.

Journal of computational biology : a journal of computational molecular cell biology·1995

Area of Science:

  • Computational neuroscience
  • Computer vision
  • Human visual perception

Background:

  • The stereo correspondence problem is fundamental to depth perception.
  • Human vision effectively solves this problem, even with complex visual input.
  • Previous models often struggle with periodic patterns and noisy image data.

Purpose of the Study:

  • To investigate how human vision solves the stereo correspondence problem using stereograms with periodic patterns.
  • To explore the role of image segmentation and planar approximations in depth perception.
  • To propose a computational model inspired by neural mechanisms in the visual cortex.

Main Methods:

  • Analysis of stereograms containing periodic patterns.
  • Modeling image segmentation based on coarse-scale features.
  • Investigating the selection of planar matches within image segments.
  • Examining the influence of feature disparity on match selection.

Main Results:

  • Stereograms are segmented by coarse-scale features, facilitating stereo matching.
  • Planar approximations are chosen within segments, suggesting a link to cortical neuron function.
  • Disparity of coarse features guides the selection of appropriate planar matches.
  • This mechanism is particularly effective for extracting reliable disparities in noisy images.

Conclusions:

  • Image segmentation and planar approximations are key components of the human stereo matching system.
  • The proposed mechanism, emphasizing planarity, likely reflects correlation-like operations in the brain.
  • This approach provides a robust method for depth estimation in challenging visual conditions, such as those with periodic patterns or noise.

Related Experiment Videos