The purpose of the proposed experiments on shape constancy and perceptual grouping is to demonstrate that perceptual grouping can both affect the results of perceptual grouping and be affected by the results of perceptual grouping. If this is in fact the case, it supports a theory in which perceptual grouping occurs at many levels in the visual system.
Experiment 1 : Perceptual grouping operates on a post-constancy representation
The purpose of this experiment is to demonstrate that perceptual grouping
can be based on a post-constancy representation. The figure at left demonstrates
this. Notice that the central column of shapes is physically identical
to those in the left three columns. However, if you perceive the central
column as printed on a surface slanted in depth, then the perceived shape
of those items should be a circle. Thus, the items would be identical to
the shapes in the right three columns assuming this post-constancy representation.
The experiment will present subjects with similar displays and ask them
with which side they perceive the central column to group the best. It
is predicted that subjects will be more likely to group the central column
with the circles (a post-constancy grouping) rather than with the physically
identical shapes.
Experiment 2 : Shape Constancy Operations are affected by the results of perceptual grouping
The purpose of this experiment is to demonstrate that shape constancy
operations operate on the results of perceptual grouping as well as influence
perceptual grouping. To do this, we will present subjects with a simple
display containing three shapes: an oval (created by drawing the projection
of a circle tilted in depth), a square, and the projection of a square
tilted into the same depth plane as the oval. These shapes are arranged
as shown in figure labeled "Neutral" below. Notice that the central oval
shape can be seen as an oval in the upright plane or as a circle tilted
in depth. The goal of the experiment is to very the degree to which the
oval groups with the upright square frame or the square tilted in depth
frame. The oval should be more likely to be perceived as having the shape
that is consistent with the frame with which it groups best. For instance,
if the oval groups most with the square tilted in depth, then it should
look be more likely to be perceived as a circle tilted in depth. This would
demonstrate that the shape constancy operations which take into account
depth information to determine true shape are sensitive to the influence
of perceptual grouping. We use the factors of similiarity of color, proximity,
and common fate motion to group the oval with the frames. Each of the factors
is demonstrated below grouping the oval with both the upright and in-depth
frames.
Neutral - oval can be seen as upright oval or circle
in-depth
|
|
|
|
|
|
 |
 |
|
|
 |
 |
|
|
 |
 |
created by Joseph Brooks, January 2002