| | |
Summary: MIT Media Laboratory Vision and Modeling Technical Report #165.
Computing Optical Flow Distributions
Using Spatio-temporal Filters
Eero P. Simoncelliy and Edward H. Adelsonz
Vision and Modeling Group, Media Laboratory
yDepartment of Electrical Engineering and Computer Science
zDepartment of Brain and Cognitive Science
Massachusetts Institute of Technology
Cambridge, Massachusetts 02139
email: eero@media-lab.media.mit.edu
November 15, 1990
Revised: March, 1991
Abstract
We describe the relationship between gradient methods for computing optical
ow and
lter-based spatio-temporal energy models of biological motion processing, revealing that
these techniques are equivalent under certain conditions. We discuss extensions of these
techniques which compute probabilitydistributions of optical
ow. The use of distributions
allows proper handling of the uncertainties inherent in the optical
ow computation, facil-
itating the combination with information from other sources. The probabilistic framework
also leads to several useful extensions of the standard quadratic gradient solution. We use
|
