Abstract:

Based on the Goldstein algorithm for InSAR interferometic phase filtering, an auto-adaptive spectrum weight function is constructed in place of the original fixed value ascertained artificially, and its exponential power parameter is calculated by the local coherence. Over filtering and under filtering problems are solved when processing the complex interferometric stripes caused by the uniform filtering. Experimental results from simulated data and ERS data indicate the validity of the modification algorithm, with good denoising effect and nice preserving of the phase resolution.

Key words: remote sensing, auto-adaptive filtering, Goldstein algorithm, interferometic synthetic aperture radar, spectrum weight function