Share Email Print

Proceedings Paper

Model-based coding for multiobject sequence
Author(s): Yunhai Liu; Lu Yu; Qingdong Yao
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

Model-based coding has been studied mostly on video sequence with only one person. Normally there are several persons in the videoconference scenes. In this paper, model-based coding for multi-objects (or multi-person) sequence is explored. A multi-scale 3D wire-frame of head has been constructed to fit for different head size on the scene. The object is composed of several components such as head, shoulder and arm. Wire- frame of shoulder and arm are built in this contribution. By connecting different components according to physical structure, a more complete wire-frame of upper body with head is built. The ways of motion parameters transferring through the different components are determined. The bit-rate of three-object sequence is compared with one-object sequence with the same image format. Model-based coding efficiency of multi-object sequence is higher than that of one-object sequence.

Paper Details

Date Published: 29 December 2000
PDF: 8 pages
Proc. SPIE 4310, Visual Communications and Image Processing 2001, (29 December 2000); doi: 10.1117/12.411866
Show Author Affiliations
Yunhai Liu, Zhejiang Univ. (China)
Lu Yu, Zhejiang Univ. (China)
Qingdong Yao, Zhejiang Univ. (China)

Published in SPIE Proceedings Vol. 4310:
Visual Communications and Image Processing 2001
Bernd Girod; Charles A. Bouman; Eckehard G. Steinbach, Editor(s)

© SPIE. Terms of Use
Back to Top