Abstract
Contour grouping in complex visual scenes requires the integration of disparate elements into a global percept, often despite the absence of local cues to grouping. One of the organizational principles used by the visual system to achieve this computationally difficult task is to perceptually group elements together when the simplest interpretation of the visual input is that the elements form a closed figure. We tested the precocity of 3 to 6 year old children with respect to their ability to use closure as an organizational principle. Both children and adults were tested on a two-alternative forced choice task in which they were asked to indicate which of two visual displays contained a contour. We manipulated both the ratio of the density of the elements on the contour to the density of the noise elements and whether the contour was open or closed. Open and closed contours differed only in terms of the orientation of a single element along the contour path (Q’s and O’s, respectively). Analysis of reaction time and accuracy data suggested that children performed near chance levels when detecting either type of contour when density of the noise was higher than the contour. For all density conditions, at least some benefit of closure was observed; children and adults were able to use closure information to improve their accuracy and reaction time. Closure became especially important for adults in terms of accuracy at higher densities, suggesting that when simpler mechanisms are sufficient relative to task demands, closure may not be recruited. This study demonstrates that children’s ability to integrate visual elements depends fundamentally on the degree to which the Gestalt comprised by those elements stands out relative to distracting cues, and that closed contours are more quickly and accurately detected than even slightly open contours.
Meeting abstract presented at VSS 2012