Polarimetric Optimisation Applied to Permanent Scatterers Identification

In this work, the potential of fully-polarimetric SAR data to enhance the performance of Permanent Scatterers Candidates (PSC) detection is investigated. In particular, the problem of finding the polarization states that maximise the signal amplitude inverse coefficient of variation (ICV) is analysed. Under the hypotheses of Rice statistics and high signal to clutter ratio (SCR), the problem can be cast into a form equivalent to the optimisation of the ratio between PS (target) and clutter backscatter. Then, solutions can be derived for the optimal transmit and receive polarisation states. In the paper, selected typologies of PS and clutter are investigated. The approach is validated through Monte Carlo simulations of the multi-temporal polarimetric response of PS-like SAR pixels. Results indicate that optimal ICV values of polarimetrically complex PSC pixels are higher than in the ERS case of the single VV channel, thus leading to increased stability of the subsequent parameter retrieval. Moreover, simulations of ICV vs. SCR suggest that a higher number of PSC may be detected by using optimal polarization states than in the conventional VV channel alone.