Abstract
The natural world provides a rich and varied array of depth information – numerous cues that are commonly divided into monocular (2D) and binocular categories. Our experience with two-dimensional representations of depth in paintings, movies and computer simulations is so compelling that it is often thought that the information supplied by the primary binocular cue, namely stereopsis, is superfluous. Stereopsis, however, provides the most precise data about the location and extent of objects along the line of sight (z-axis). Most of the monocular cues to depth, such as perspective, relative size and texture, depend on changes in angular subtense associated with increasing distance. The minimal detectable change (1 – 2%) in angular subtense requires a change in distance that is generally substantially larger than the change in distance associated with the smallest detectable difference in disparity, at least at moderate distances (<15 meters). Although this conclusion is based on geometry, I will present experimental evidence showing that binocular depth thresholds measured with real objects in a cluttered environment are as much as 10 times better than monocular thresholds. The precision of stereopsis is also important in guiding hand movements, particularly grasp. Thus, the 2 – 3% of the population that are stereo-blind due to crossed eyes (strabismus) in childhood suffer a significant disadvantage in both depth judgments and eye-hand coordination. Recent studies (Ding & Levi, 2011) have shown that with practice some adult strabismics can recover some degree of stereopsis, raising the hope that better therapies in strabismic children can restore this important dimension.