Abstract
The spatial correlation properties of the backscattered power in polarimetric SAR data are characterized by means of a two-point statistic provided by a wavelet frame that acts as a differential operator. In particular, the dependence of this statistic on a rotation of the linear polarization basis is considered. On the one hand, a theoretical model is proposed, which enables the numerical simulation of the wavelet variance of the crosspol and copol power synthesized in the rotated basis, given the power spectra of the feature vector in the H,V basis. On the other hand, experimental analysis of polarimetric SAR data is undertaken using wavelet variance estimates as a function of scale and polarization state. Results are encapsulated in a graphical form called the wavelet scaling polarimetric signature (WASPS).
Original language | English |
---|---|
Pages (from-to) | 1433-1436 |
Journal | International Geoscience and Remote Sensing Symposium |
DOIs | |
Publication status | Published - 12 Nov 2012 |
Keywords
- wavelet
- two-point statistic
- SAR
- polarimetry