The Discriminability of Local Surface Structure

Abstract
The ability of observers to discriminate depth and orientation differences between separated local regions on object surfaces was examined. The objects were defined by many optical sources of information simultaneously, including shading, texture, motion, and binocular disparity. Despite the full-cue nature of the displays, the observers' performance was relatively poor, with Weber fractions ranging from 10% to 40%. The Weber fractions were considerably lower for discriminations of surface-orientation differences than for similar discriminations of depth differences. The ability of observers to discriminate surface-orientation differences was approximately invariant over the separation of the regions in the projected image. In contrast, the ability to discriminate depth differences was highly influenced by the amount of image separation. This qualitative difference between the perception of depth intervals and surface-orientation differences suggests that knowledge of depths and orientations may be represented separately within the human visual system.