ABSTRACT
Coherent change detection (CCD) use the sample degree of coherence as a measure of the temporal change collected between two complex-valued SAR images observed in the same area and using also the same geometry and polarization. The problem of classical CCD approaches shows a temporal change only when there are in areas of the image that have high clutter-to-noise power ratio. All experiments regarding the maximum-likelihood (ML) CCD, found to estimate useful information also in lower clutter-to-noise power ratio. Experiments used only electromagnetic transmissions in the same polarization. This work extends the formulation of the probability distribution function of the CCD in a multi-dimensional version useful for multi-polarization SAR data. Results of this ML-polarimetric interferometric SAR-CCD (ML-PolInSAR-CCD) show surprising recovery of both amplitude and phase CCD information. This information recovery is useful for improved interferometric SAR (InSAR), permanent scatterers interferometry (PS-InSAR) and SAR tomography applications.
Disclosure statement
No potential conflict of interest was reported by the author.