New Methods in Multibaseline Polarimetric SAR Interferometry Coherence Optimization

A new extension method in the equal scattering mechanism (ESM) from single baseline to multibaseline (MB) for polarimetric synthetic aperture radar interferometry (PolInSAR) coherence optimization is proposed in this letter. However, despite this, this new method and the traditional available methods such as Best, ESM, MB-ESM, and suboptimum scattering mechanism have their disadvantages in the framework of differential interferometric SAR (DInSAR) applications. The ESM method cannot guarantee the global maximum in theory and just optimizes an approximation formula of the original coherence definition, which leads to a deviation. The method using exhaustive search polarimetric optimization (ESPO) must search four parameters one by one in a defined step size and cost the main computational drawback. Focusing on the disadvantage of these methods, this letter proposes another new method to transform the 4-D optimization problem into two independent 2-D problems, which costs less time than that by ESPO with basically the same accuracy. Seven full polarimetric RADARSAT-2 images are taken for experiment, and the results show that this new method has a better effect in computation time and accuracy for DInSAR applications.