Share Email Print

Proceedings Paper

New real-time algorithm for multiview disparity-based 3D-object segmentation
Author(s): Yuecheng Zhang; Yun He
Format Member Price Non-Member Price
PDF $14.40 $18.00
cover GOOD NEWS! Your organization subscribes to the SPIE Digital Library. You may be able to download this paper for free. Check Access

Paper Abstract

With the development of video communication system, more and more attention has been put on the implementation of varieties of functionality associated with content. Thus, the object segmentation algorithms features both semantic and automatic are tremendous necessary. In this paper a new real-time algorithm for 3D-object segmentation algorithm based on disparity is proposed. Its kernel disparity matching methodology is developed on base of a general pixel-based segmentation algorithm known a Dynamic Programming on Disparity Space Image. With the introduction of weighted feature span, the computational complexity is reduced apparently. The proposed algorithm is tested on both real captured stereo sequence and synthetic multiview sequence, and gets satisfying result in generality, quality and complexity.

Paper Details

Date Published: 29 December 2000
PDF: 11 pages
Proc. SPIE 4310, Visual Communications and Image Processing 2001, (29 December 2000); doi: 10.1117/12.411850
Show Author Affiliations
Yuecheng Zhang, Tsinghua Univ. (China)
Yun He, Tsinghua 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