Speckle reduction of SAR images using wavelet-domain hidden Markov models

Wavelet-domain hidden Markov models (HMMs), proposed bu M. S. Crouse et al. (1998), are used for speckle reduction of SAR images. The method is a frameworks for statistical signal processing and is based on HMM and wavelets. The HMM is a tree-structured probabilistic graph that captures the statistical properties of the coefficients of the wavelet transform. Both wavelet and translation-invariant wavelet denoising based on HMMs are studied. Results on denoising of SAR images are presented. The proposed method shows great promise for speckle removal and hence provides good detection performance for SAR based recognition.