Optic Flow

Optic flow is the pattern of apparent motion of surfaces, edges, and texture in the visual field generated by an observer's movement through an environment. It is not a feature of the world itself but a structured transformation of the optic array produced by the observer's own locomotion. Psychologist James J. Gibson argued that optic flow is the fundamental information for visual perception: it specifies not only the direction of self-motion but the layout of the environment, the distance of surfaces, and the time-to-contact with approaching obstacles.

The key insight is that optic flow is lawful. It is not random visual noise to be filtered by the brain but a mathematically structured gradient that directly specifies the three-dimensional structure of the environment relative to the moving observer. The focus of expansion in the flow pattern indicates the direction of heading; the rate of flow specifies distance; the differential motion between foreground and background specifies depth. The perceiver does not compute these properties from the flow; the perceiver detects them as invariant structures within the flow itself.

This makes optic flow a central concept in embodied cognition and robotics, where it provides a principled alternative to reconstructive approaches that build depth maps from multiple camera views. The optic flow is already the depth information, specified in a form that a moving observer can use directly.

The computational vision community's insistence on treating optic flow as a "problem to be solved" — a noisy input from which depth must be inferred — reveals the depth of the representationalist prejudice. Optic flow is not a problem; it is the solution that evolution found.