Abstract
Purpose: :
Stereothresholds measured with local-contour stereograms are typically lower than those measured with complex random-dot stereograms (RDSs). These dissimilar thresholds may be due to differences in the characteristics of local and global stereograms, e.g., spatial frequency content, contrast, inter-element separation, or area. Because the density of stimulus elements generally differs between local and global stereograms, the goal of this study was to compare stereothresholds for a short line segment presented against static and dynamic RD backgrounds of various densities.
Methods: :
The stimuli were RDSs programmed in MATLAB with the psychtoolbox extension. Each dot was convolved with a 2D Gaussian distribution to allow subpixel resolution. Four corrected observers with normal binocular vision judged the depth of a line (0.85’ x 11’) presented near the center of each 171’ x 171’ stereogram. Dot density in the stereogram ranged from 0.75% (4 dots) to 15% (841 dots). Stimuli were presented in a dark room at a distance of 114 cm using a mirror haploscope. The target line and background dots were matched in detectablity for each RDS density by presenting the stimuli at a fixed multiple of the contrast-detection threshold. Stereothresholds were defined as the semi-interquartile range of the cumulative Gaussian function, fit to the percentage of near responses as a function of the line disparity.
Results: :
Stereothresholds average approximately 12" for a static RD density of 2.7%, and 14" for a dynamic RD density of 1.77%, and approximately double for both lower and higher densities.
Conclusions: :
Based on previous results, the increase in stereothresholds as dot density increases is not due to a change in the number of RDs. Different processes are likely to increase stereothresholds for low vs. high RD background density. Specifically, the increase in stereothresholds at low densities is likely due to the increased spacing between elements, which reduces their usefulness as a reference for relative disparity judgments. The increase in stereothresholds at high densities is likely to result from a crowding effect. However, similar stereothresholds can be obtained for RD backgrounds of low and high density.
Keywords: binocular vision/stereopsis • depth • perception