A new method for speckle reduction in Synthetic Aperture Radar (SAR) images using optimal window size

9th Symposium of the International Society for Digital Earth

地球科学;计算机科学

Mahdavi, S.^1 ; Salehi, B.^1,2 ; Moloney, C.^1 ; Huang, W.^1 ; Brisco, B.^3

Department of Electrical and Computer Engineering, Memorial University of Newfoundland, NL, Canada^1

C-CORE/LOOKNorth, St. John's

NL, Canada^2

Canada Center for Mapping and Earth Observation, Ottawa

ON, Canada^3

关键词: Adaptive window size;    Minimum Mean Square Error (MMSE);    Newfoundland , Canada;    Optimal window size;    Radiometric quality;    Real and imaginary;    Synthetic aperture radar (SAR) images;    Visual interpretation;   

Others  :  https://iopscience.iop.org/article/10.1088/1755-1315/34/1/012021/pdf
DOI  :  10.1088/1755-1315/34/1/012021

学科分类：计算机科学（综合）

来源: IOP

【 摘 要 】

Speckle degrades the radiometric quality of a Synthetic Aperture Radar (SAR) image and makes its visual interpretation difficult. The approaches proposed previously for speckle filtering of SAR images exploit a window of fixed size for this purpose. But a fixed size window is not sufficient as the size of objects may vary throughout the image. In this paper, a method is introduced by which each pixel in the image is filtered using a window size which is optimal for that pixel. Real and imaginary parts of a single-channel SAR image are used for the selection of the best window size for each pixel, and then intensity image is filtered by applying that window size. The Average and Minimum Mean Square Error (MMSE) filters are modified using the Adaptive Window Size method. This approach is implemented on the HH-channel of a RADARSAT-2 image acquired over the Avalon Peninsula near St. John's, Newfoundland, Canada. This filter can supress speckle effectively while retaining the details reasonably.