Classification of urban areas in multi-date ERS-1 images using structural features and a neural network

19 November 2002

conference paper
Published by Institute of Electrical and Electronics Engineers (IEEE)

Vol. 2, 901-903 vol.2
https://doi.org/10.1109/igarss.1995.521091

Abstract

Describes a new method to extract structural informations from images. The loss of spatial resolution and distortions-from edges, as it occurs with standard texture algorithms, are reduced to a minimum. Furthermore, the authors describe the inhomogeneity by three different structure types according to the structures contained in SAR images. Finally they use a neural network (RBF-Network) to get a more precise classification of urban areas from SAR images.

Keywords

This publication has 6 references indexed in Scilit:

Efficient speckle filtering of SAR images
Published by Institute of Electrical and Electronics Engineers (IEEE) ,2002
Approximation capabilities of multilayer feedforward networks
Neural Networks, 1991
Layered Neural Networks with Gaussian Hidden Units as Universal Approximations
Neural Computation, 1990
Fast Learning in Networks of Locally-Tuned Processing Units
Neural Computation, 1989
Multivariate Generalizations of the Wald-Wolfowitz and Smirnov Two-Sample Tests
The Annals of Statistics, 1979
Statistical and structural approaches to texture
Proceedings of the IEEE, 1979