September 2005
Volume 5, Issue 8
Vision Sciences Society Annual Meeting Abstract  |   September 2005
Contour shape effects on search performance: evidence for constant curvature coding
Author Affiliations
  • Patrick Garrigan
    University of California, Los Angeles
  • Philip J. Kellman
    University of California, Los Angeles
Journal of Vision September 2005, Vol.5, 468. doi:
  • Views
  • Share
  • Tools
    • Alerts
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Patrick Garrigan, Philip J. Kellman; Contour shape effects on search performance: evidence for constant curvature coding. Journal of Vision 2005;5(8):468. doi:

      Download citation file:

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

  • Supplements

In natural scenes, physical contours are not always visible along their entire extent. Often, spatially disparate parts of the contour must be linked to establish shape. Field, Hayes, & Hess (1993) studied path-linking among separated Gabor elements and found that detection of the path depended on the spacing, orientation, and alignment of the constituent elements. Later, Yen & Finkel (1998) modeled contour integration using a grouping mechanism that considered interactions among pairs of oriented units. Effects of higher-order relations, such as curvature polarity, have also been found (Pettet, 1999).

We hypothesized that neural circuits involved with shape representation might facilitate detection, via reduction of target uncertainty. Specifically, we conjectured that oriented segments related by constant curvature play a role in shape representation and thereby produce certain efficiencies in detection.

We report three experiments using a modified version of the paradigm developed by Field, et al. (1993). The results show that detection of a path defined by four oriented elements was enhanced if the orientations of the elements were constrained to lie along a path of constant curvature. This result cannot be explained by pairwise interactions among oriented units and suggests sensitivity to properties more specific than that demonstrated for consistent curvature polarity reported earlier (Pettet, 1999). We also found performance differences between constant curvature as compared to varying turn-angle targets in a priming / detection paradigm.

Taken together, these experiments suggest an interaction between neural shape representation mechanisms utilizing constant curvature, contour interpolation, and search and detection performance. We simulated these results using a modified version of the model developed by Yen & Finkel (1998) that incorporates higher-order interactions among oriented units, providing added salience to constant curvature contours.

Garrigan, P. Kellman, P. J. (2005). Contour shape effects on search performance: evidence for constant curvature coding [Abstract]. Journal of Vision, 5(8):468, 468a,, doi:10.1167/5.8.468. [CrossRef]
 Supported by NEI EY13518 to PJK

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.