Abstract
The studies of the Gestalt principles in the literature focused on the spatial arrange of stimulus component and often neglects the effect of luminance contrast. However, luminance contrast may play a major role in perceptual grouping (Wilson & Wilkinson, 1998; Lin et al., 2017). The current study investigated contrast dependency of Gestalt proximity law with a tripole Glass patterns (tGPs) paradigm. A tGPs consists of randomly-distributed sets of three dots, or tripole. Each tripole contained one anchor dot and two context dots. An observer would perceive a clockwise (CW) or counter-clockwise (CCW) spiral global percept by grouping the anchor dot with either one of the context dot. We changed luminance contrast of the context dots ranging between -30 and 0 dB, while the luminance contrast for the anchor dot was fixed at -15 dB. The anchor-context distance varied between 2.5 and 20 min. Participants were asked to report whether they perceived CW or CCW in the tGP in each trial. When the luminance contrast of both context dots was low, the probability of perceiving a CW spiral increased when the distance between CW context dot and the anchor dot decreased, as would be predicted by the proximity grouping principle. However, when the luminance contrast was high, the probability of perceiving a CW spiral decreased as the distance between CW context and anchor dot decreased. This result was in opposite from the prediction from the proximity law. These results suggested that the Gestalt proximity law is subject to contrast control. Our result can be well fit by a divisive inhibition model in which the response for a global pattern is the summed excitation of linear filters, each for a local dipoles, raised by a power and then divided by a divisive inhibition signals from all other dipoles.
Meeting abstract presented at VSS 2018