A new class of information about depth across an edge and figure-ground organization -- edge texture grouping -- is described and demonstrated. The central issue in both cases is whether the edge between two regions "belongs to" (or "is grouped with") one side or the other. The side that is grouped with the edge should be perceived as closer and shaped by the edge, whereas the ungrouped side should be perceived as farther away and extending (unshaped) behind the edge. This analysis implies that classic grouping factors that relate various edge properties to corresponding textural properties within adjacent regions should systematically influence perceived depth across an edge and figural status. We report the results of several experiments that strongly support this claim for the grouping factors of common fate, proximity, synchrony, and similarity of blur, color, and orientation. Further, we argue that evidence of grouping in certain moving and flickering displays that are ecologically unnatural suggest that these effects are mediated by grouping processes rather than by inferences based on simple ecological statistics. The edge-texture grouping hypothesis provides a unified account of the present phenomena and others previously reported in the depth perception literature (e.g., Yonas, Craton & Thompson, 1987) as well as a coherent explanation of Weisstein's anomalous findings about the effects of spatial frequency and flicker on depth and figural status (e.g., Klymenko & Weisstein, 1986). Edge-texture grouping effects have important implications for depth and shape algorithms in computer vision as well as for corresponding perceptual and neural processes in human vision.
Common Fate Motion
The displays below demonstrate that when the luminance edge is grouped
with a textured region by the common motion of the edge and the texture
elements of that region, the region is overwhelmingly seen as figural.
Three conditions are shown. Part A shows the edge moving in common fate
with the texture elements of the left region. As expected, the left region
looks figural in this condition. Condition B shows that the effect is not
due merely to the motion of the dots. In fact, in condition B it is the
static region that is seen as figural. This is because the texture elements
of that region (the right region in this case) are in static common fate
with the edge while the dots of the other region (the left) are moving.
Condition C simply shows that the effect still occurs in displays in which
both regions of texture elements are moving. The left region is seen as
figural because its texture elements are moving with the edge.
Wong and Weisstein (1987) demonstrated that regions with texture elements that flickered were more likely to be seen as background. This effect is shown in stimulus A below. The right region looks like background because it is flickering. At face value, this finding is curious because there is no clear rationale, ecological or otherwise, why flickering regions should be seen as behind non-flickering regions. However, this phenomenon is a clear prediction of the edge-texture grouping hypothesis. Because the edge groups with the non-flickering region by virtue of being static, they belong together and that region is seen as figural. This suggests a clear prediction not tested by Wong and Weisstein. What happens when the edge and a region flicker together. According to the Wong and Weisstein hypothesis that flickering regions are more likely to be seen as ground, the flickering region should still be seen as ground. The edge-texture grouping hypothesis predicts the opposite. Stimulus B below demonstrates that the figure-ground organization is consistent with the predictions of the edge-texture grouping hypothesis. Stimulus B demonstrates the effects of both flicker grouping as well as synchrony. Stimulus C shows both the edge and the right texture elements flickering. However, the flicker is out of phase. The strength of the figure-ground organization is much more ambiguous here. However, participants are still much more likely to judge the right side as figural. Overall, flicker similarity and synchrony grouping between and edge and a textured region influence figure-ground organization.
|Stimulus A||Stimulus B||Stimulus C|
Weisstein demonstrated that the relatively higher spatial-frequency region will tend to be seen as figural. They used two adjacent regions like in the examples below. As texture in the region, they used spatial frequency gratings. Stimulus A shows an example of this principle using the dot textures that we have employed above. The left side should have more higher spatial frequency components than the right side because the right side elements have been blurred. The edge-texture grouping hypothesis explains this result. Because the edge in stimulus A is a sharp edge, it should group best with the sharp dots on the left side and make this side figural. Furthermore, the edge-texture grouping hypothesis predicts that the blurred (or lower spatial frequency) region should be figural when the edge is also blurred. Stimulus B is an example of this condition. We have found that participants are significantly more likely to see the blurred side as figural when the edge is also blurry.
|Stimulus A||Stimulus B|
Notice that in the columns labeled "left" and "right", the central curvy
edge has the same color as the texture elements of one of the adjacent
regions. The hypothesis is that the edge will group more with that region
and thus be more likely to be seen as figural. For instance, in the upper
left most box, the left side texture elements are the same as the edge.
This side should be more likely to be seen as figural in brief displays.
The "Same Color" and "Neither" conditions are included to show that when
no the edge groups with neither side by color similiarity, there is no
bias to choose one side over the other.
In this example, proximity of the texture dots to the edge is hypothesized to group the edge with the region and thus lead to that region as figural. For instance, in the upper-left most box, the texture dots on the left side of the edge are closest to the edge. Thus, the left region should group with the edge by proximity and be more likely to be seen as figural than the right region. This was done for both line edges (row 1) as well as luminance edges (rows 2 and 3). The "Middle" and "Outside" conditions demonstrate that there is no bias in figure-ground organization when the dots are equally close to the edge on both sides.
In this example, similarity between the orientations of the texture elements and the orientation of the elements that make up the edge serves as the grouping factor. The top row contains images in which the edge is composed of diagonally-oriented segments. Thus, the side containing diagonally-oriented texture elements should group by similarity with the edge and be more likely to be seen as figural. The left column should be biased to have the left side as figural, the central column neither and the right column has the right side biased to be figural. The bottom row demonstrates the same phenomenon except that the edges are now composed of horizontally and vertically oriented components and thus regions with horizontally and vertically oriented texture elements should group and be likely to be seen as figural.
by Joseph Brooks, updated December 2003