Interferometric SAR Phase Unwrapping Based on Markov Random Fields

摘要

For Digital Elevation Model (DEM) generation using InSAR technique, phase unwrapping is always a crucialrnand sophisticated issue. In this paper, a new phase unwrapping algorithm based on Markov random fields isrnproposed. Taking priori knowledge of original phases into account, a Markov random field is used to formulaternthe relationship between wrapped phases and their original values. By analysis on the Markov random field, it isrnconvenient to illustrate the local properties of the original phases. In the end, experiments on synthetic and realrndata prove the validation of this algorithm.