Summary
2011 International Symposium on Nonlinear Theory and Its Applications
2011
Session Number:A3L-A
Session:
Number:A3L-A1
Identification of Number of Target Image Regions Based on Bifurcation of Fixed Point in Discrete-Time Coupled Neuronal System
Ken’ichi Fujimoto, Mio Kobayashi, Tetsuya Yoshinaga, Kazuyuki Aihara,
pp.156-159
Publication Date:2011/9/4
Online ISSN:2188-5079
Summary:
We developed a discrete-time coupled neuronal system that can extract image regions in a given image and exhibit them in a time series. We call this function "dynamic image segmentation", and it is performed by synchronizing the oscillatory responses from discrete-time neurons. In this paper, we describe how the topological property of a fixed point in the system gives significant information on the structure of regions in a given image. We also suggest a novel way to identify the number of image regions without performing a segmentation process.