Spectral Pattern Recognition by a Two-Layer Perceptron: Effects of Training Set Size

Fischer, Manfred M. and Staufer-Steinnocher, Petra (1996) Spectral Pattern Recognition by a Two-Layer Perceptron: Effects of Training Set Size. Discussion Papers of the Institute for Economic Geography and GIScience, 55/96. WU Vienna University of Economics and Business, Vienna.


Download (6MB)


Pattern recognition in urban areas is one of the most challenging issues in classifying satellite remote sensing data. Parametric pixel-by-pixel classification algorithms tend to perform poorly in this context. This is because urban areas comprise a complex spatial assemblage of disparate land cover types - including built structures, numerous vegetation types, bare soil and water bodies. Thus, there is a need for more powerful spectral pattern recognition techniques, utilizing pixel-by-pixel spectral information as the basis for automated urban land cover detection. This paper adopts the multi-layer perceptron classifier suggested and implemented in [5]. The objective of this study is to analyse the performance and stability of this classifier - trained and tested for supervised classification (8 a priori given land use classes) of a Landsat-5 TM image (270 x 360 pixels) from the city of Vienna and its northern surroundings - along with varying the training data set in the single-training-site case. The performance is measured in terms of total classification, map user's and map producer's accuracies. In addition, the stability with initial parameter conditions, classification error matrices, and error curves are analysed in some detail. (authors' abstract)

Item Type: Paper
Keywords: Two-layer perceptron / multispectral landcover data / supervised pattern classification / training set size / stability and parameter conditions
Divisions: Departments > Sozioökonomie > Wirtschaftsgeographie und Geoinformatik > Fischer
Depositing User: ePub Administrator
Date Deposited: 26 May 2014 11:28
Last Modified: 17 Jul 2015 10:56
URI: https://epub.wu.ac.at/id/eprint/4160


View Item View Item


Downloads per month over past year

View more statistics