Purchase this article with an account.
Andrew Meso, Johannes Zanker; On the contribution of form and motion cues in the perception of transparency. Journal of Vision 2007;7(9):403. doi: 10.1167/7.9.403.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Transparency is the perceptual separation of a region of visual field into two or more depth layers based on available visual cues. Interpreting motion and form (luminance and shape) cues is crucial for biological systems to make sense of the complex visual environments which they inhabit. Form and motion induced transparency have predominantly been investigated as separate phenomena. We combine them, using transparently moving stimuli in which form cues can be manipulated. We designed a 2 interval 2AFC task in which the perceived strength of transparently moving test stimuli is measured by repeated, randomly interleaved comparisons with reference stimuli which form a measurement scale. One such scale was made up of a pair of moving luminance gratings in which the number of odd Fourier harmonics used to make the gratings forms a measure of perceived transparency strength. Speed difference, contrast and spatial frequency are kept constant in the reference. Paired luminance gratings moving in the same direction at different speeds can be perceived as transparent, but to a lesser degree than gratings moving in opposite directions. Hence, modulation by form cues becomes measurable under this ‘same direction’ condition. Using a computational model of global motion based on correlation motion detectors, we predict component separability in these test stimuli from the distribution of their pooled local motion signals, while varying spatial frequency and amplitude parameters of one of the paired gratings. Results from presenting the same stimuli to participants using the 2AFC task deviate from predictions based only on motion cues. Lowering amplitude of one grating in the test pair unexpectedly increases transparency strength. Increasing spatial frequency differences between the grating pairs also unexpectedly decreases transparency strength. We discuss these results in the context of a general transparency percept, which integrates all available cues that reveal differences between the two components.
This PDF is available to Subscribers Only