Abstract
The perceived depth from disparity in random-dot stereogram depends on both luminance contrast and the spatial frequency (SF) of depth modulation (Chen et al., 2016). Here, we varied the number of cycles to investigate whether this effect may be due to local disparity gradient or global surface configuration. The test stimuli were rectangular random dot stereograms (1.27 × 3.44 degree) with three kinds of surface configuration: (a) a low SF (0.29 cy/deg) single-cycle cosine bulge, (b) a high SF (0.87 cy/deg) single-cycle cosine bulge, and (c) a high SF (0.87 cy/deg) three-cycle corrugated surface. For each surface configuration, maximum test disparity ranged from −20 to 20 arcmin while luminance contrast ranged from 5% to 80%. The observers adjusted the length of a horizontal bar to match the perceived depth in the test stimuli. In all conditions, the perceived depth for both near and far disparities was an inverted-U shape matching function of physical disparity. As in our previous study, both the perceived depth magnitude and the disparity of the peak of the matching function increased sigmoidally with luminance contrast. Both the magnitude and peak disparity were the largest for high SF single-cycle bulge followed by the low SF single-cycle bulge, and then the three-cycle corrugated surface. Our results suggest that not only local disparity gradient but also global surface configuration affects disparity processing. Such effects can be captured by a multiple-stage model in which the perceived depth is determined by the weighted average of several nonlinear contrast gain control mechanisms, each with a different disparity selectivity. Furthermore, since more than one disparity lies in the channel receptive field, the response of the cells vary with disparity gradient, hence, generating different perceived depth for different surface configurations.
Acknowledgement: MOST(Taiwan) 107-2420-H-002-029-DR