Rotation-Invariant Image Association for Endoscopic Positional Identification Using Complex-Valued Associative Memories
In order to solve the problem for identifying the tip position of fiber scopes under the endoscopic surgery, we propose an image association system. The proposed system is realized by combining complex-valued associative memories with a 2-dimensional discrete Fourier transform, and it is capable of rotation-invariant image association. Simulation results using surface images observed through endoscope are also presented.
Unable to display preview. Download preview PDF.
- 1.Eiju Watanabe: Endoscopic surgery guided by neuronavigator, Clinical Neuroscience (in Japanese), Vol.16,No.12, pp.1419–1422 (1988)Google Scholar
- 3.Igor Aizenberg, Naum. Aizenberg, Joos Vandewalle: Multi-Valued and Universal Binary Neurons. Kluwer Academic Publishers (2000)Google Scholar
- 4.A. Hirose: Proposal of fully complex-valued neural networks, Proc. of IJCNN’92 Baltimore, U.S.A, vol.IV, pp.152–157, (1992).Google Scholar
- 5.H Aoki: Analysis of equilibrium states of hopfield associative memory extended to complex number field. IEICE Trans., vol.J78-A,no.9, pp.1238–1241, (Sept. 1995)Google Scholar
- 7.H. Aoki, MR. Azimi-Sadjadi, Y. Kosugi: Image Association Using a Complex-Valued Associative Memory Model. IEICE Trans. Fundamentals, Vol.E83-A,No.9, pp.1824–1832, (Sept. 2000).Google Scholar
- 8.H. Aoki, Y. Kosugi: An Image Storage System Using Complex-Valued Associative Memories. 15th International Conference on Pattern Recognition Barcelona 2000 Vol. 2 pp.626–629 (Sept. 2000)Google Scholar