Motion-Disparity Interaction and the Scaling of Stereoscopic Disparity

Contemporary studies of visual perception often view the observer's problem as one of estimation. In other words, the observer seeks to estimate various aspects of the scene, such as the size and shape of the objects in that scene. The information available to inform this estimation is viewed as consisting of one or more visual cues. Each cue may be used on its own to estimate some aspects of scene geometry. Most cues require additional information (visual or otherwise) for the information from that cue to be fully interpreted. This chapter concentrates on the visual cue to depth of binocular disparity. The cue is the disparity in location of features in the two eyes. Horizontal disparity, the difference in horizontal position in each of the two eyes' views of objects in a scene, is a powerful visual cue to the three-dimensional structure of the world. However, horizontal disparities cannot be fully interpreted without some knowledge of the viewing geometry (the locations, gaze directions, and torsional states of the two eyes). For central gaze (fixating straight ahead), their interpretation requires an estimate of the distance to the fixation point (the fixation distance). Most cues to depth require estimates of the viewing geometry to produce metric estimates of depth in a scene. In a model of depth-cue combination (Landy et al., 1995), this process is described as cue promotion. Cues are promoted by the insertion of the values of the unknown viewing geometry and resolution of depth ambiguities. Without promoting the cues, their raw data (e.g., disparities and velocities) are in different units so that simple cue-combination strategies, such as averaging the depth estimates made using each cue, are impossible. When the missing parameters are the eye positions (vergence, gaze directions, and torsions), the promotion process is referred to as depth scaling. In particular, in central gaze, the raw sensory data for the cue (velocities, disparities, etc.) are scaled by (that is, multiplied by, or multiplied by the square of) an estimate of the fixation distance. To the extent that this scaling is done accurately, the result is depth constancy: perceived depth that is independent of changes in viewing conditions. In this hap-ter we will limit our discussion of cue promotion to the issue of scaling by the fixation distance. We review a number of ways in which depth scaling may be accomplished.