Terrain misclassification problem – analysis using pattern simulation approach

TitleTerrain misclassification problem – analysis using pattern simulation approach
Publication TypeBook Chapter
Year of Publication2015
AuthorsJasiewicz, Jarosław, Alfred Stach, and Jakub Nowosad
Secondary AuthorsJasiewicz, Jaroslaw, Zbigniew Zwoliński, Helena Mitasova, and Tomislav Hengl
Book TitleGeomorphometry for Geosciences
Pagination153 - 156
PublisherBogucki Wydawnictwo Naukowe, Adam Mickiewicz University in Poznań - Institute of Geoecology and Geoinformation
CityPoznań, Poland
ISBN Number978-83-7986-059-3
Abstract

We present the results of a numerical experiment aiming at explaining reasons for classification errors when using an automatic pattern-based terrain classifications algorithm proposed by Jasiewicz et. al. [3]. We use composition of landform elements from incorrectly classified areas, and we use texture pattern from example areas to synthesize a new “terrain” which inherits properties from both sources. Using a new Pattern Analysis Toolbox (GeoPAT, [4,5]) we found that classification errors come from convergence of landscape properties: after replacing texture in misclassified areas with texture as indicated by an example area a new synthetic area shows higher degree of similarity to the landscape class from which it inherits texture. It allow to draw conclusion that short-range textural properties is  that feature which at that moment best describes diversity of landscapes for automatic classifications.

AttachmentSizeHitsLast download
Jasiewicz2015geomorphometry.pdf598.89 KB7153 hours 56 min ago
3.142855
Your rating: None Average: 3.1 (7 votes)