Automatic computation of speckle standard deviation in SAR images

Being a coherent reception system, Synthetic Aperture Radar (SAR) sensors are highly liable to speckle noise effect, which masks details and patterns in the image, and therefore, degrades interpretation. Speckling may be reduced by applying filtering techniques to SAR multilook images. The major pro...

Descripción completa

Guardado en:

Detalles Bibliográficos
Publicado:	2000
Materias:	filter image analysis synthetic aperture radar
Acceso en línea:	https://bibliotecadigital.exactas.uba.ar/collection/paper/document/paper_01431161_v21_n15_p2883_Frulla http://hdl.handle.net/20.500.12110/paper_01431161_v21_n15_p2883_Frulla
Aporte de:	Biblioteca Digital - Facultad de Ciencias Exactas y Naturales (UBA) de Universidad de Buenos Aires

id	paper:paper_01431161_v21_n15_p2883_Frulla
record_format	dspace
spelling	paper:paper_01431161_v21_n15_p2883_Frulla2023-06-08T15:11:45Z Automatic computation of speckle standard deviation in SAR images filter image analysis synthetic aperture radar Being a coherent reception system, Synthetic Aperture Radar (SAR) sensors are highly liable to speckle noise effect, which masks details and patterns in the image, and therefore, degrades interpretation. Speckling may be reduced by applying filtering techniques to SAR multilook images. The major problem that arises from this type of method is the estimation of input parameters: sliding window size and speckle standard deviation. The present paper describes a manual and two automatic methods devised to estimate speckle standard deviation based on the texture concept, in order to extract homogenous regions. The automatic method were specially developed to improve results obtained with the manual one, and the so-called least-squares approach and mean approach were considered. The mean approach was introduced as an alternative to the least-squares approach. It performs better in terms of computing time and disk space use, and even shows a slightly higher accuracy when tested against artificially speckled images. Manual and automatic methods were applied as an example using ERS-1/SAR one-look and three-look images with different features, obtained over several Austral and Antartic regions of Argentina. Results show that the automatic method is a valuable tool for estimating speckle standard deviation, being accurate, less tedious, and preventing typical human errors associated with manual tasks. © 2000 Taylor & Francis Group, LLC. 2000 https://bibliotecadigital.exactas.uba.ar/collection/paper/document/paper_01431161_v21_n15_p2883_Frulla http://hdl.handle.net/20.500.12110/paper_01431161_v21_n15_p2883_Frulla
institution	Universidad de Buenos Aires
institution_str	I-28
repository_str	R-134
collection	Biblioteca Digital - Facultad de Ciencias Exactas y Naturales (UBA)
topic	filter image analysis synthetic aperture radar
spellingShingle	filter image analysis synthetic aperture radar Automatic computation of speckle standard deviation in SAR images
topic_facet	filter image analysis synthetic aperture radar
description	Being a coherent reception system, Synthetic Aperture Radar (SAR) sensors are highly liable to speckle noise effect, which masks details and patterns in the image, and therefore, degrades interpretation. Speckling may be reduced by applying filtering techniques to SAR multilook images. The major problem that arises from this type of method is the estimation of input parameters: sliding window size and speckle standard deviation. The present paper describes a manual and two automatic methods devised to estimate speckle standard deviation based on the texture concept, in order to extract homogenous regions. The automatic method were specially developed to improve results obtained with the manual one, and the so-called least-squares approach and mean approach were considered. The mean approach was introduced as an alternative to the least-squares approach. It performs better in terms of computing time and disk space use, and even shows a slightly higher accuracy when tested against artificially speckled images. Manual and automatic methods were applied as an example using ERS-1/SAR one-look and three-look images with different features, obtained over several Austral and Antartic regions of Argentina. Results show that the automatic method is a valuable tool for estimating speckle standard deviation, being accurate, less tedious, and preventing typical human errors associated with manual tasks. © 2000 Taylor & Francis Group, LLC.
title	Automatic computation of speckle standard deviation in SAR images
title_short	Automatic computation of speckle standard deviation in SAR images
title_full	Automatic computation of speckle standard deviation in SAR images
title_fullStr	Automatic computation of speckle standard deviation in SAR images
title_full_unstemmed	Automatic computation of speckle standard deviation in SAR images
title_sort	automatic computation of speckle standard deviation in sar images
publishDate	2000
url	https://bibliotecadigital.exactas.uba.ar/collection/paper/document/paper_01431161_v21_n15_p2883_Frulla http://hdl.handle.net/20.500.12110/paper_01431161_v21_n15_p2883_Frulla
_version_	1768543794108563456

Automatic computation of speckle standard deviation in SAR images

Ejemplares similares