Share Email Print

Proceedings Paper

Dynamic search of Gaussian segmentation
Author(s): Shun Ido; Satoshi Arai; Ryo Takamatsu; Makoto Sato
Format Member Price Non-Member Price
PDF $17.00 $21.00
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

This paper presents a new approach, called Gaussian segmentation, to extract regions containing locally concentrated stimuli. We describe two kinds of searching methods, local search and dynamic search, to link local solutions of the estimating function in parameter space in the same and different visual fields, respectively. In the local search, we propose an iterative technique which can find a local solution from any initial state of parameters by means of searching the parametric space locally. The dynamic search is provided to move from one of the local search results as the initial state to the other local search results. Finally, the availability of the technique is shown by experiments for estimating using functional images and real images.

Paper Details

Date Published: 30 September 1996
PDF: 8 pages
Proc. SPIE 2826, Vision Geometry V, (30 September 1996); doi: 10.1117/12.251806
Show Author Affiliations
Shun Ido, Tokyo Institute of Technology (Japan)
Satoshi Arai, Tokyo Institute of Technology (Japan)
Ryo Takamatsu, Tokyo Institute of Technology (Japan)
Makoto Sato, Tokyo Institute of Technology (Japan)

Published in SPIE Proceedings Vol. 2826:
Vision Geometry V
Robert A. Melter; Angela Y. Wu; Longin Jan Latecki, Editor(s)

© SPIE. Terms of Use
Back to Top