Contents of this page are obsolete. This page is preserved and stored at this URL just from historical viewpoint. Original URL was http://www.mm.media.kyoto-u.ac.jp/members/kameda/...
Please visit www.kameda-lab.org for recent information. (2002/12/06, [email protected])

Publication, Research, Docs and Info, KAMEDA


next up previous
Next: Introduction

Three Dimensional Motion Estimation of a Human Body Using a Difference Image Sequence

KAMEDA Yoshinari, MINOH Michihiko, IKEDA Katsuo

Abstract:

We propose a model-matching method to estimate a motion of an human body from a difference image sequence. We assume that shape information of the human body and camera calibration parameters have been given to the system in advance. In the algorithm, parts which consists of the human body are classified into 3 modes: moving, stationary, and occlusion mode. When a part is in moving mode, the system matches the part's projection with a difference image. when it is in stationary mode, the system doesn't use the image. when in occlusion mode, the system predicts the motion from its locus and doesn't use the image.





Yoshinari Kameda
Mon Apr 7 14:56:47 JST 1997