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

Psychophysics with junctions in real images.

Josh McDermott1

  • 1Perceptual Science Group, Massachusetts Institute of Technology, NE20-444, 3 Cambridge Center, Cambridge, MA 02139, USA. jhm@mit.edu

Perception
|November 25, 2004
PubMed
Summary

Human vision relies on junctions, but detecting them in real images requires larger areas than previously thought. Many junctions need complex processing, suggesting they are identified after initial scene interpretation.

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

Perceptual spaces: mathematical structures to neural mechanisms.

The Journal of neuroscience : the official journal of the Society for Neuroscience·2013
Same author

The evolution of music.

Nature·2008
Same author

Nonhuman primates prefer slow tempos but dislike music overall.

Cognition·2006
Same author

Probing the evolutionary origins of music perception.

Annals of the New York Academy of Sciences·2006
Same author

The geometry of the occluding contour and its effect on motion interpretation.

Journal of vision·2004
Same author

Are consonant intervals music to their ears? Spontaneous acoustic preferences in a nonhuman primate.

Cognition·2004

Area of Science:

  • Vision Science
  • Computational Neuroscience
  • Image Processing

Background:

  • Junctions, where image contours meet, are crucial for early visual processing.
  • They are considered simple, local features providing information about occlusions and transparency.
  • The detectability of local junctions with early visual mechanisms is debated.

Purpose of the Study:

  • To investigate if locally defined junctions in real images are detectable by simple, early visual mechanisms.
  • To quantify the spatial extent required for human observers to reliably identify occlusion junctions.
  • To determine the role of local versus global processing in junction detection.

Main Methods:

  • Human observers labeled occlusion points in real images.
  • A second group judged the presence of occlusion junctions in variable-sized circular image regions.
  • Control experiments addressed multi-scale effects.

Main Results:

  • Observer performance in identifying occlusion junctions was poor for small regions (under 50 pixels).
  • Reliable detection required significantly larger regions than predicted by simple local mechanisms.
  • Control experiments confirmed the findings were not scale-dependent.

Conclusions:

  • While some image junctions are locally defined and detectable by early vision, many require more complex, global processing.
  • The detection of a substantial fraction of junctions may occur after initial scene interpretation.
  • Rethinking the role of simple, early mechanisms in junction processing is necessary.

Related Experiment Videos