A New Alternative for SAR Imagery Coherence Estimation

Nowadays, multidimensional Synthetic Aperture Radar (SAR) imagery has raised as a powerful tool to extract physical information about the Earth surface. One of the main parameters which characterizes these multidimensional data is the complex correlation coefficient between two SAR images. It has been demonstrated, in the related literature, that this parameter can be sensible to different physical properties depending on the nature of the multidimensional SAR system. Hence, a correct estimation of it is essential. This paper focus on the estimation of the amplitude of this correlation coefficient, called coherence. A novel approach for its estimation is presented in the following, which does not present the drawbacks of classical approaches as the topography bias in interferometric applications or the loss of spatial resolution linked to the estimation due to the spatial filtering.