Title :
PolSAR image speckle reduction based on polarimetric decomposition and classification
Author :
Han, Ping ; Dong, Fei ; Wu, Renbiao
Author_Institution :
Tianjin Key Lab. for Adv. Signal Process., Civil Aviation Univ. of China, Tianjin, China
Abstract :
On account of the inherent speckle in PolSAR (Polarimetric Synthetic Aperture Radar) images and target scattering properties maintaining after speckle suppressing, an effective algorithm for speckle reduction based on scattering model is developed. First, fast alternative to H/α is used in PolSAR image decomposition, and total back-scattering power is combined for initial p ixel classification. Then, minimum distance measure and h ierarchical cluster method are used for further classification. Finally, pixels are filtered according to their classes. Experimental results with AIRSAR data show that, the new algorithm is more effective than Scattering-Model-Based speckle filter developed by Lee not only in speckle reduction but also in polarimetric properties preservation.
Keywords :
filtering theory; image classification; image denoising; pattern clustering; radar imaging; radar polarimetry; synthetic aperture radar; AIRSAR data; PolSAR image speckle reduction; hierarchical cluster method; initial pixel classification; minimum distance measure; polarimetric classification; polarimetric decomposition; polarimetric properties preservation; polarimetric synthetic aperture radar images; scattering-model-based speckle filter; target scattering properties; total backscattering power; Classification algorithms; Filtering; Filtering algorithms; Matrix decomposition; Oceans; Scattering; Speckle; PolSAR image; classification; fast alternative to H/α; polarimetric decomposition; speckle reduction;
Conference_Titel :
Radar (Radar), 2011 IEEE CIE International Conference on
Conference_Location :
Chengdu
Print_ISBN :
978-1-4244-8444-7
DOI :
10.1109/CIE-Radar.2011.6159651
