December 2001
Volume 1, Issue 3
Free
Vision Sciences Society Annual Meeting Abstract  |   December 2001
Contour interpolation: Necessary but not sufficient for the perception of interpolated contours
Author Affiliations
  • S. E. Guttman
    University of California, Los Angeles, CA, USA
  • P. J. Kellman
    University of California, Los Angeles, CA, USA
Journal of Vision December 2001, Vol.1, 384. doi:10.1167/1.3.384
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to Subscribers Only
      Sign In or Create an Account ×
    • Get Citation

      S. E. Guttman, P. J. Kellman; Contour interpolation: Necessary but not sufficient for the perception of interpolated contours. Journal of Vision 2001;1(3):384. doi: 10.1167/1.3.384.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

PURPOSE: How do we perceive complete objects from fragmented visual information? We recently reported evidence for the operation of low-level contour interpolation mechanisms between appropriately oriented, outlined, notched-circle inducers (ARVO 2000). These data fit with theories suggesting that contour completion processes “fill in” missing contours whenever visible edges satisfy geometric constraints that would allow the interpolation of a smooth, monotonic contour. However, outline inducers fail to produce a subjective experience of illusory contours. Here, we investigate further how low-level contour interpolation mechanisms operate on edges that are relatable, but do not give rise to illusory or amodal contour perception. METHODS: Observers viewed shapes defined by filled or outline notched-circle inducers, plus-shaped inducers, partly occluded edges, and corner elements. Observers performed a fat-thin classification task that could be carried out either by judging the overall shape depicted by grouping the elements or by examining the orientations of individual elements. In separate experiments, we measured both reaction times in speeded classification tasks, and discrimination performance (d′) for briefly presented shapes. RESULTS: Classification performance was both accurate and rapid when the visible edges satisfied certain spatial relations, regardless of the nature of the inducing elements. However, manipulations that disrupted the geometry of edge relatability, such as misalignment, turning inducers outward, or eliminating tangent discontinuities, dramatically reduced performance. CONCLUSIONS: The data indicate that low-level contour interpolation mechanisms operate between appropriately oriented edges, even for stimuli in which no interpolated contours are perceived. Conscious perception of a modal or amodal contour may be blocked at a later stage if boundary assignment is inconsistent with the existence of a central, occluding shape.

Guttman, S.E., Kellman, P.J.(2001). Contour interpolation: Necessary but not sufficient for the perception of interpolated contours [Abstract]. Journal of Vision, 1( 3): 384, 384a, http://journalofvision.org/1/3/384/, doi:10.1167/1.3.384. [CrossRef]
Footnotes
 Supported by NSF LIS-9720410 to PJK, and an NSERC postgraduate scholarship to SEG.
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×