The role of surface completion on the convexity context effect

In order to represent and interact with our environment, the visual system has to perceptually organize the retinal image into potential objects and their spatial relations. One aspect of perceptual organization is figure-ground segregation, the process of identifying which parts of a scene are figure and which are background. One tendency is for the visual system to assign convex regions as figure and concave regions as ground. Recently, this convexity bias was discovered to increase when the number of repeating figure-ground regions increases. It has been hypothesized that this convexity context effect (CCE) is caused by observers perceptually completing the concave regions into a single background behind the convex figures. If true, then the CCE should occur even when regions are made discontinuous by another surface that partially occludes them but should not occur when regions are discontinuous with no occluding surface to explain the discontinuity. The results of my project partially confirm these predictions in that partially occluded displays produced an equal magnitude CCE as unoccluded displays. However, discontinuous displays without an occluding surface also produced an equal magnitude CCE as unoccluded contiguous displays. Follow-up experiments will address this failure of our control condition.