| | |
Summary: M.I.T. Media Laboratory Vision and Modeling Group, Technical Report No. 221, April 1993
Appears in Proceedings of the IEEE Computer Vision and Pattern Recognition Conference,
pp. 361{366, New York, June 1993.
Layered Representation for Motion Analysis
John Y. A. Wang Edward H. Adelson
The MIT Media Laboratory The MIT Media Laboratory
Dept. Elec. Eng. and Comp. Sci. Dept. Brain and Cognitive Sciences
Massachusetts Institute of Technology Massachusetts Institute of Technology
Cambridge, MA 02139 Cambridge, MA 02139
Abstract
Standard approaches to motion analysis assume
that the optic
ow is smooth; such techniques have
trouble dealing with occlusion boundaries. The most
popular solution is to allow discontinuities in the
ow
eld, imposing the smoothness constraint in a piece-
wise fashion. But there is a sense in which the dis-
continuities in
ow are artifactual, resulting from the
attempt to capture the motion of multiple overlapping
objects in a single
ow eld. Instead we can decom-
pose the image sequence into a set of overlapping lay-
|
