The Community for Technology Leaders
Proceedings of 1994 28th Asilomar Conference on Signals, Systems and Computers (1994)
Pacific Grove, CA, USA
Oct. 31, 1994 to Nov. 2, 1994
ISSN: 1058-6393
ISBN: 0-8186-6405-3
pp: 1362-1366
Qi Li , Dept. of Electr. Eng., Rhode Island Univ., Kingston, RI, USA
D.W. Tufts , Dept. of Electr. Eng., Rhode Island Univ., Kingston, RI, USA
R.J. Duhaime , Dept. of Electr. Eng., Rhode Island Univ., Kingston, RI, USA
P.V. August , Dept. of Electr. Eng., Rhode Island Univ., Kingston, RI, USA
ABSTRACT
Two methods of classification and related fast training algorithms are compared with each other and with backpropagation in this paper. The first method is the discriminant neural network (DNN). One hidden node is added at each design stage until the DNN meets the design requirements. The second method uses the radial basis function network (RBF). We modify the RBF by solving a succession of binary classification problems in order to provide fast training. These two classification methods are applied to automatically classify 14 categories of land cover using multispectral aerial images. We find that the training times for the DNN and the modified RBF (MRBF) are much less than the training times for backpropagation or RBF. The performances of DNN (72%) and MRBF (60%) are better than obtained by linear discriminant analysis (LDA) (55%). The resulting structure and computations are simpler for the DNN than for the other methods.<>
INDEX TERMS
image classification, backpropagation, feedforward neural nets, spectral analysis
CITATION

Qi Li, D. Tufts, R. Duhaime and P. August, "Fast training algorithms for large data sets with application to classification of multispectral images," Proceedings of 1994 28th Asilomar Conference on Signals, Systems and Computers(ACSSC), Pacific Grove, CA, USA, 1995, pp. 1362-1366.
doi:10.1109/ACSSC.1994.471680
96 ms
(Ver 3.3 (11022016))