Possiblity of surface matching using Hopfield's neural network

• Autorzy: Ke L. C.

Warianty tytułu

PL

Możliwość zastosowania sieci neuronowej Hopfield’a do spasowania powierzchni terenu

• Języki publikacji: EN

Abstrakty

EN

In the last decade the usefulness of neural networks (NN) in engineering and commerce has been widely developed. On the satellite images the different tasks such as cloud, ice, sea-surface thermal feature tracking, spatial resolution improvement, pattern classification etc. have been applied, using neuron networks. In digital photogrammetry neural networks have been used, at first, in solution of stereo vision problem called stereo matching and to other tasks such as detection, location of object from, mobile mapping data and another. This paper presents an idea of using Hopfield's neural network (HNN) to solve the surface matching problem. first of all the energy function (or cost function) of surface matching would be provided. for this purpose all constraint conditions with the target (or goal) of surface matching will be combined into energy function, which can be applied by Hopfield's neural network in determining the optimal surface matching problem.

PL

Przy porównaniu dwóch powierzchni terenu, które są reprezentowane dwoma zbiorami danych punktów otrzymanych z różnych sensorów tego samego obszaru wykorzystano sieć neuronową Hopfiel'da HNN dla uzyskania rozwiązania optymalnego, tzn. kiedy funkcja energii tej sieci zdąża do minimum. Funkcję te wyznaczono z sumy warunków zapewniających jednoznaczność cech należących do odpowiednich elementów powierzchni oraz z funkcji celu. Stabilny stan sieci HNN jest określony wówczas, gdy funkcja energii dąży do minimum. W takim przypadku zadanie dopasowania dwóch powierzchni jest najlepsze.

Słowa kluczowe

EN

neural networks; surface area; optimization theory

PL

sieci neuronowe; powierzchnia terenu; teoria optymalizacji

• Wydawca: Komitet Geodezji PAN
• Czasopismo: Geodezja i Kartografia
• Rocznik: 2003
• Tom: T. 52, z. 2
• Strony: 71--83
• Opis fizyczny: Bibliogr. 16 poz., rys., wykr.

Twórcy

autor

Ke L. C.

• Warsaw University of Technology Institute of Photogrammetry and Cartography

Bibliografia

• [1] Cote S., Tatnall A.R.T., The Hopfield neural network as a tool for feature tracking and recognition from satellite sensor image. IEEE Transaction on Geosciences and Remote Sensing. Vol. 18, p. 871-885, No 4, Match 1997.
• [2] Habib A., Kelly D., Asmamaw A., New approach to solving matching problems in photogrammetry. IAPRS, Vol. 33, Part B2, p. 257-264, Amsterdam 2000.
• [3] Hong Fan, Zuxun Zheng, Daosheng Du: A Hopfield neural network algorithm for automated name placement for point feature. IAPRS, Vol. XXXIII, part B4, p. 262-268. Amsterdam 2000.
• [4] Hopfield J.,Tank W., Neural computation of decision in optimization problem. Biological cybernetics. Vol. 52, p. 141-152, Jan. 1985.
• [5] Li Rongxing, Wang Weian, Tseng Hong-Zeng, Detection and location of objects from mobile mapping image sequences by Hopfield neural network. PE&RS. Vol. 65, No.10, p. 1199-1205. Oct. 1999.
• [6] Loung Goung, Tam Zheng, Stereo matching using artificial neural networks. IAPRS Vol. 2, part B3, p.417-421. Washington 1992.
• [7] Maria del Carmel Valdes, Minoru Inamura, Spatial resolution improvement of remotely sensor image by a fully interconnected neural network approach. IEEE Transaction on Geosciences and Remote Sensing. Vol. 38, No 5, p. 2426-2430. Sept. 2000.
• [8] Osowski S., Sieci neuronowe w ująciu algorytmicznym (Neural networks in algorithms). WNT, Warszawa 1996.
• [9] Rutkowska D., Piliński M., Rutkowski L., Sieci neuronowe, algorytmy genetyczne i systemy rozmyte (Neural networks, genetic algorithms and fuzzy systems). PWN, Warszawa 2000.
• [10] Schenk T., Krupnik A., Postolov J., Comparative study of surface matching algorithms. IAPRS. Vol. 33, part B4, p. 518-524. Amsterdam 2000.
• [11] Tatem Andrew J., Lewis Hugh G., Peter M. Atkingson, Mark S. Mixon, Super revolution target identification from remotely sensed images using a Hopfield neural network. IEEE Transaction on Geosciences and Remote Sensing. Vol. 39, No 4, p. 781-796. April 2001.
• [12] Tseng Yi Hsing, Orienting digital stereo pairs by matching Fourier descriptors. IAPRS. Vol. XXXI, part B3, p. 880-885. Vienna 1996.
• [13] Wierzbicki A., Findeisen W., Szymanowski J., Teoria i metody obliczeniowe optymalizacji (Theory and coputation methods of optimization). PWN, Warszawa 1977.
• [14] Jnwook Go, Gunhee Han, Hagbae Kim, Multigradient: A new neural network learning algorithm for pattern classification. IEEE Transaction on Geosciences and Remote Sensing. Vol. 39, p. 986-993. No 5, May 2001.
• [15] Palma Blonda, Andrea Baraldi, Elisabetta Binaghi, Comparison of the multiplayer perceptron with neuron-fuzzy techniques in the estimation of cover class mixture in remotely sensed data. IEEE Transaction on Geosciences and Remote Sensing. Vol. 39, p. 994-1005. No 5, May 2001.
• [16] Yin-Cheng Lee, Chin-Teng Lin, Her-Chang Pu, Satellite sensor image classification using cascaded architecture of neural fuzzy network. IEEE Transaction on Geosciences and Remote Sensing. Vol. 38, p. 1033-1043. No 2, March 2001.