A Novel Coherence Estimation Method for InSAR

The coherence between Synthetic Aperture Radar (SAR) single-look complex(SLC) images, as an indicator of the similarity of Interferometric pair and the quality of Interferometric phase, has been widely applied for exploratory activities. However, coherence estimation always results in unsatisfactory accuracy due to non-ideal factors, such as nonstationary signals and estimation bias. A novel coherence estimation method is proposed to improve its accuracy for interferometric SAR (InSAR). The proposed method is modified from three aspects: the selection of statistical homogeneous points based on the coherence matrix, compensation for the interferometric phase based on a nonlinear phase model, and bias mitigation based on a mean filter. Results of both simulated and real data demonstrate that in comparison with the existing methods, the proposed method achieves higher reliability and clearer spatial features in the coherence maps, especially for complex scenes.