ABSTRACT
Eye Tap video is a new genre of video imaging facilitated by and for the apparatus of the author’s eyeglass-based “wearable computer” invention [1]. This invention gives rise to a new genre of video that is best processed and compressed by way of comparametric equations, and comparametric image processing. These new methods are based on an Edgertonian philosophy, in sharp departure from the traditional Nyquist philosophy of signal processing. A new technique is given for estimating the comparameters (relative parameters between successive frames of an image sequence) taken with a camera (or Eye Tap device) that is free to pan, tilt, rotate about its optical axis, and zoom. This technique solves the problem for two cases of static scenes: images taken from the same location of an arbitrary 3-D scene and images taken from arbitrary locations of a flat scene, where it is assumed that the gaze pattern of the eye sweeps on a much faster time scale than the movement of the body (e.g., an assumption that image flow across the retina induced by change in eye location is small compared to that induced by gaze pattern).
