The study of interferogram denoising method based on EMD and adaptive filter

In view of different characteristics in different intrinsic mode function (IMFS) after signal and noise were decomposed by empirical mode decomposition (EMD), a new adaptive filter method based on empirical mode decomposition for reducing interferometric phase noises was proposed. At first the signals were decomposed by empirical mode decomposition, and different high-frequency IMF signals were separately processed by adaptive filter with different filter gradient parameter, and then the scale information corresponding to the noises was subtracted from the original interferogram so as to achieve the purpose of noise suppression, at last the results were compared with the five methods of Goldstein filter, periodic pivoting median filter, EMD decomposition method, the adaptive filter and Baran filter. The experiment results show that EMD-adaptive filter method suppresses effectively the InSAR interferogram phase noises, and can well keep the information of the phase details and the edges.