Purchase this article with an account.
Huseyin Boyaci, Fang Fang, Scott O. Murray, Daniel Kersten; Amodal completion affects lightness perception. Journal of Vision 2007;7(9):235. doi: https://doi.org/10.1167/7.9.235.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
It is well-known that perceptual grouping influences the perception of surface lightness; however the extent of the interaction between grouping and lightness is not known. Amodal completion - when two spatially separated surfaces appear to be grouped together behind an occluder - is one of the mechanisms which can facilitate perceptual grouping. We study perceptual grouping and its effect on lightness perception through amodal completion.
Imagine quantizing the luminances in the rectangular form of the classic “Craik-O'Brien-Cornsweet” stimulus to three levels, resulting in four uniformly gray rectangular regions. Like the “Craik-O'Brien-Cornsweet” stimulus, the leftmost and rightmost regions (flanks) have the same luminance. The quantization eliminates the luminance gradients in the bands on either side of the central border, but introduces two additional vertical borders. If these two borders are occluded by two bars, then the surface appears amodally completed and one perceives the leftmost and rightmost rectangular regions to differ in lightness, similar to the “Craik-O'Brien-Cornsweet” effect: the flank, which is spatially closer to the higher luminance band, is perceived to have a higher lightness relative to the other flank.
We studied the magnitude of the effect as a function of the contrast of the central border in the stimulus described above. Three observers viewed the 3-D stimuli binocularly through a stereoscope and performed a pixel matching task. We found that the lightness effect initially gets stronger as the contrast of the central border increases reaching a maximum, then weakens and nearly disappears as the contrast approaches 100 percent. Preliminary fMRI data show that activity in early visual cortical areas reflects the illusory percept. Candidate computational models that can potentially predict the perceptual effect are examined.
This PDF is available to Subscribers Only