Paper Info
Title
Local texture stationarity indicator for filtering DoÑana wetlands SAR images
Abstract
This paper defines a new operator, named Ds, for local texture stationarity assessment on SAR images.. The aim is to discriminate heterogeneous targets from land cover types of high normalized variance values, as those observed in flooded vegetation areas of Doñana wetlands. Suitable Ds thresholds for such discrimination were estimated for different window sizes through Monte Carlo simulations of synthetic textures. Maximum stationary texture windows were then determined on Doñana ASAR scenes by Ds multi-resolution thresholding and averaging was applied within. Results reveal the substantial degree of smoothing achieved over high variance cover types, while edges among different targets were properly preserved.
Year
DOI
Venue
2012
10.1109/IGARSS.2012.6352513
Geoscience and Remote Sensing Symposium
Keywords
Field
DocType
floods,geophysical image processing,hydrological techniques,radar imaging,synthetic aperture radar,Donana ASAR scenes,Donana wetlands SAR images,Ds multiresolution thresholding,Monte Carlo simulations,flooded vegetation,heterogeneous targets,high normalized variance values,land cover,local texture stationarity assessment,local texture stationarity indicator,maximum stationary texture windows,suitable Ds thresholds,synthetic textures,SAR,Texture stationarity,multi-resolution filtering,wetlands
Computer vision,Radar imaging,Vegetation,Normalization (statistics),Computer science,Synthetic aperture radar,Remote sensing,Filter (signal processing),Smoothing,Artificial intelligence,Thresholding,Land cover
Conference
ISSN
ISBN
Citations 
2153-6996 E-ISBN : 978-1-4673-1158-8
978-1-4673-1158-8
1
PageRank 
References 
Authors
0.36
3
6
Name
Order
Citations
PageRank
Belen Marti-Cardona110.70
Carlos López-Martínez232.77
Josep Dolz-Ripolles310.70
Marti-Cardona, B.410.36
Lopez-Martinez, C.552.86
Dolz-Ripolles, J.610.36