Unsupervised model based SAR data denoising
Dušan Gleich, Mihai P. Datcu · 2005
In this paper a wavelet based method for SAR data denoising is presented. The SAR image is corrupted by a multiplicative noise that can be modeled as an additive noise in wavelet domain. In this paper an image is modeled as a Gauss Markov random field and noise is considered as Gaussian with unknown variance. An unsupervised stochastic model based approach to image denoising is presented. The parameters are estimated from incomplete data using mixtures of wavelet coefficients, and expectation maximization algorithm. Observed wavelet coefficient is estimated using inter and intra scale of wavelet coefficients to estimate image and noise model parameters. Presented wavelet based method efficiently removes noise from SAR images.