Visual observers are aware of the structure of their surrounding space. For the stationary observer this optical space differs systematically from physical space. The deviations are indeed surprisingly large. Most of our present understanding has been gained under extremely reduced experimental conditions, such as an evironment reduced to a small number of dim point sources in otherwise complete darkness, fixed head and body posi- tions, sometimes even restrictions on eye movements. In such circumstances observers are forced to rely on binocular disparity only. Although one might expect simple results in such cases, there is actually no satisfactory consensus in the literature as to the structure of human optical space.

We propose to deviate from this tradition, in that we will use a structured environment and allow the observer free head movements. Moreover, whereas the literature focusses on the metrical structure, we propose to address the more fundamental projective structure of optical space. There exist sufficient indications from the literature that a satisfactory metric model of optical space might be utopic. We have found indications that its projective structure might be quite tight though. Because we address the projective rather than the metric structure, we need novel psychophysical methods, such as exocentric pointing and collinearity or coplanarity judgements. In pilot experiments we have found these methods to be viable and effective.

The goal is to arrive at a coherent geometrical model of human optical space, possibly not fully metrical and quite likely dependent upon the physical structure of the observer's environment.