Directionally adaptive despeckling of SAR image using interscale dependence

Directionlet Transform (DT) has gained popularity over the last few years as an anisotropic, critically sampled and perfect reconstruction transform with directional vanishing moments along any two directions. The performance of despeckling schemes based on multi-resolution analysis can be improved significantly by taking into account the multi-scale correlation among the transform coefficients. This paper proposes a novel directionally adaptive de-speckling algorithm for SAR images by taking into account the statistical inter scale dependency of Cauchy-Gaussian modeled DT coefficients. The effectiveness of the proposed scheme is illustrated by comparing it with other similar schemes.