Abstract:

This paper presents a new method for identification and selection of persistent scatterer (PS) pixels in SAR images based on wavelet phase analysis. The wavelet filtering theory is introduced into the PS pixel identification and selection. The differential phase interferograms are filtered using the interferogram filtering method based on the wavelet phase analysis. Noise phase and its indicator, and probability of candidate PS pixels are then estimated. Taking into account stability of the amplitude and phase of the candidate PS pixels, the PS pixels are effectively identified and selected from the SAR image. Theoretical analysis and experimental results show that the PS pixels identified and selected using the proposed method are effective and reliable.

Key words: persistent scatterer (PS) pixels, wavelet phase analysis, noise phase indicator, PS probability