IMAGE FLOW THEORY: A FRAMEWORK FOR 3-D INFERENCE FROM TIME-VARYING IMAGERY
ALLEN M. WAXMAN Boston University
KWANGYOEN WOHN University of Pennsylvania
A major source of three-dimensional information about objects in the world is available to the observer in the form of time-varying imagery. Relative motion between textured objects and observer generates a time- varying optic array at the image, from which image motion of contours, edge fragments and feature points can be extracted. These dynamic features serve to sample the underlying "image flow" field. We present an overview of our work on image flow theory and its potential for the recovery of three-dimensional surface structure and space motions of rigid objects. Methods for flow recovery and segmentation stem directly from the analytic structure of image flow fields. New, closed-form solutions are given for the structure and motion of planar and curved surface patches from monocular image flow. The relationship between instabilities in the computation and visual illusions are also explored. Finally, the fusion of motion analysis and stereo vision is discussed in the context of "binocular image flows."
Ever since the pioneering experiments of Wallach and O'Connell ( 1953), Johannson ( 1975), and Braunstein ( 1976, 1983) made clear that an object's three-dimensional (3-D) structure could be revealed through two- dimensional (2-D) image motion, the significance of time-varying imagery in the perceptual process has become well established. The early
Questia, a part of Gale, Cengage Learning. www.questia.com
Publication information: Book title: Advances in Computer Vision. Volume: 1. Contributors: Christopher Brown - Editor. Publisher: Lawrence Erlbaum Associates. Place of publication: Hillside, NJ. Publication year: 1988. Page number: 165.
This material is protected by copyright and, with the exception of fair use, may not be further copied, distributed or transmitted in any form or by any means.