Abstract
Previous studies have shown that oppositely directed motions enhance the perception of depth from disparity when the different directions are spatially intermixed as in a simulated rotating transparent cylinder defined by random dots. One argument for this enhancement is that opposite directions of motion help to solve the stereo correspondence problem. Is this enhancement observed when the opposite directions of motion are spatially segregated reducing or eliminating the ambiguity in stereo correspondence? Observers judged a step-change in depth in a small three-panel display with the motion direction in the middle panel opposite to that in the upper and lower panels (standard) against the depth in a simultaneously available display with the directions of motion the same in the three panels (test). Each panel was a window which contained a drifting grating of moderate contrast providing distinct features that should have simplified the solution of the correspondence problem. The disparity in the test display was varied to produce a forced-choice psychometric function from which we could estimate the point-of-subjective-equality at which the depths in the test and standard displays looked the same. For three observers, the PSE's indicated greatly enhanced depth in the standard display relative to the test display. Observers also judged the depths in both of these displays against a display with binocular disparity but no motion, and the results were consistent; displays with relative motion in opposite directions coupled with disparity always appeared to have more depth than displays with no relative motion coupled with similar amounts of disparity. Conclusion: the enhancement of stereoscopic depth provided by relative motion is not necessarily related its role in solving the stereo correspondence problem because it is also observed when the opposite directions are spatially segregated.