The physical logic of visual perception

How does the visual system resolve the ambiguities inherent in the retinal data to achieve a veridical representation of the natural, 3D environment? The problem was discerned and addressed by the Gestalt psychologists, who developed evidence for a simplicity principle underlying the emergence of the phenomenal image. Many past investigators have attempted to concretize this principle based on maximizing simplicity of geometrical description, without arriving at a general rule or algorithm. Here, I attempt to demonstrate that the visual representation is derived on the basis of constraints that mediate (a) the transformation of the 2D outline into a 3D shape, and (b) phenomenal features of this shape pertaining to viewer-object interaction. These constraints reflect (a) physical laws; (b) the typical structure of the environment and the objects in it; and (c) the typical behavior of the human viewer. More specifically, the process is based on an implicit knowledge of the tendency for the shape of 3D objects to mediate their possible equilibrium states in a ground/gravity environment---a fact which is used by the process to determine the direction of gravity; a tendency to assume maximum whole-symmetry of the 3D volume inferred from the 2D form; assumptions about the likely viewpoints a particular organism will adopt relative to ground/gravity; a tendency to use the shape of the 2D form when it is centered on the retina for the purpose of resolving it as a 3D form. In addition, the orientation of the phenomenal object is shown to be represented such as to facilitate motor interaction with it by the organism. The special problems of pictorial perception are also addressed, and shown to be consistent, albeit it indirectly, with ordinary perception.