Dynamical systems approach to binocular stereopsis
Altman, Edward James
This item is only available for download by members of the University of Illinois community. Students, faculty, and staff at the U of I may log in with your NetID and password to view the item. If you are trying to access an Illinois-restricted dissertation or thesis, you can request a copy through your library's Inter-Library Loan office or purchase a copy directly from ProQuest.
Permalink
https://hdl.handle.net/2142/22874
Description
Title
Dynamical systems approach to binocular stereopsis
Author(s)
Altman, Edward James
Issue Date
1990
Doctoral Committee Chair(s)
Ahuja, Narendra
Department of Study
Electrical and Computer Engineering
Discipline
Electrical and Computer Engineering
Degree Granting Institution
University of Illinois at Urbana-Champaign
Degree Name
Ph.D.
Degree Level
Dissertation
Keyword(s)
Engineering, Electronics and Electrical
Artificial Intelligence
Language
eng
Abstract
The extraction of depth in binocular stereopsis relies upon the ability to detect the co-occurrence of similar features in a stereo pair of images and the subsequent extraction of positional disparity. The fundamental problem of stereopsis is to achieve coordinated activity through the linkage of co-occurring features. This thesis investigates a model for stereopsis in which the linkage of features is achieved by interactions among low-order dynamical systems serving as feature detectors. In this model, a dynamical system is an ensemble of nonlinear oscillators associated with an image location. The initial state of each oscillator is determined by the gray-level context of the image location. Distinct features can be identified through the convergence of the dynamical system to a stable mode of coherent oscillations. The recovery of binocular disparity is achieved by the interactions between two such dynamical systems associated with corresponding epipolar lines in the two images.
The dynamical systems examined in this thesis have multiple stable states, which can be used for pattern classification and exhibit behavior similar to that of neural networks. Our approach differs from that of neural networks in that the stable states of the system are characterized by limit cycle oscillations rather than stable fixed points. In the course of this thesis, it is shown that the nonlinearities of the dynamical systems play a key role in maintaining coherent oscillations among the coupled elements, as well as establishing stable interactions between dynamical systems responding to the co-occurrence of similar features in the stereo pair of images.
Use this login method if you
don't
have an
@illinois.edu
email address.
(Oops, I do have one)
IDEALS migrated to a new platform on June 23, 2022. If you created
your account prior to this date, you will have to reset your password
using the forgot-password link below.
