SAR Image Despeckling Using Local Contextual Hidden Markov Model in the Contourlet Domain

Synthetic Aperture Radar (SAR) image despeckling is an important problem in the SAR applications. A novel despeckling approach using a local contextual hidden Markov model (LCHMM) in the contourlet domain is presented in this paper. The proposed method can not only use the multiresolution and multidirection characteristics of the contourlet transform, but also exploit the local statistics and capture the intrascale dependencies of the contourlet coefficients by using LCHMM. The experiments in despeckling SAR images show that the proposed method in contrary to other methods can obtain a better trade-off between smoothing the homogeneous areas and keeping the edges and can get better visual effect.