ML-Based Fringe-Frequency Estimation for InSAR

This letter focuses on estimating the local fringe frequency of the interferometric phase, under the hypothesis of superficial scattering. Starting from the formulation of the maximum-likelihood estimator, a new simplified estimator is derived. Due to computational efficiency and robustness versus model errors, the resulting estimator is suited for large data processing in the presence of model uncertainty. Furthermore, such an estimator can be straightforwardly extended to the multi-baseline case, resulting in the possibility to estimate the terrain slope with great accuracy. An application to real data is presented, based on a multi-baseline ENVISAT data set.