Abstract
The role of this study was to assess the role of binocular cues in the perception of 3D shape under natural viewing conditions. While studies of binocular vision often use simple stimuli in order to isolate binocular cues, in the natural world there are many, highly redundant sources of information regarding the 3D structure of the environment. It is important to establish the role of binocular cues in the perception of depth in this natural context. Binocular natural scenes were captured under controlled viewing conditions using two digital cameras. Participants were presented with monocular and binocular images in a dark room using a standard Wheatstone stereoscope at a distance of 50cm. Following Koenderink (1996), observers were shown a number of pairs of dots on the surface at different locations and were asked to discriminate the depth of these dots, i.e. to assess which point was the closer of the two. The dots were distinguished by their structure as a disk and a circle. Observers' responses were converted into depth inequalities and were used to determine (i) relief maps of perceived depth and (ii) the internal consistency of observers' local depth judgments. Contrary to the expectation from using simple stimuli, the pictorial relief maps could not be meaningfully generated for both viewing conditions. Observers' responses show good repeatability of responses to individual pairs of points. Responses were more consistent under monocular viewing. The current study shows little indication of the contribution of binocular cues to this process under natural viewing conditions.