AN N-DIMENSIONAL RECURRENT NOISY PHASE UNWRAPPING ALGORITHM

摘要

Phase unwrapping arises as a key step in several imaging technologies, from which we emphasize interferometric synthetic aperture radar (InSAR), synthetic aperture sonar, magnetic resonance imaging. Due to various decorrelation factors, such as spatial, temporal, spectral, thermal noise, and so on, there is much phase noise in InSAR interferogram. The phase noise degrades the visual quality of the interferogram, increases the errors of the results derived from the InSAR interferogram, and obstructs the phase unwrapping. In addition to accuracy, computational efficiency is very important, recurrent phase unwrapping method appears to be one of the most computational efficient methods proposed for this problem. However, its beaver in presence of noise is unknown. In this paper, one-dimensional (1-D) recurrent phase unwrapping algorithm is used to eliminate phase discontinuities in N-dimensional (N-D) complex arrays. The behavior of this method in the presence of different noise levels is studied and the problem of phase filtration and its effect on recurrent algorithm behavior is considered. Tests of the method presented illustrate the validity and effectiveness of proposed algorithm for difficult problems with noisy and discontinuous original phases.