Research of Different Quality Map Applying in Kalman Filter Phase Unwrapping

Abstract

The original Kalman filter phase unwrapping method uses the coherence obtained from the coherence map to determine the observation equation noise variance. If the coherence map is not obtained, we can use three kinds of quality map to determine the observation equation noise variance. This paper respectively uses pseudo-coherence, phase derivative variance and maximum phase gradient obtained from the pseudo-correlation map, phase derivative variance map and maximum phase gradient map to determine the noise variance. We apply Kalman filter phase unwrapping method for the interferometry synthetic aperture radar data in the two different conditions: the steep terrain and the flat terrain. And we analyze and compare the results in two aspects: the visual analysis and the quantitative analysis. The experiment results show that the result using the phase derivative variance to Kalman filter phase unwrapping is most reliable in the steep terrain, the result using the pseudo-correlation map as the quality map is most precise in the flat terrain.