Anais > Edições Anteriores > Anais do XIX WCF 2023 > A Proposition of a Bayesian Filter for Despeckling SAR Images Using Conjugate Distribution, Stochastic Distances, and Non-Local Means

A Proposition of a Bayesian Filter for Despeckling SAR Images Using Conjugate Distribution, Stochastic Distances, and Non-Local Means

Autores: Pedro H. S. A. Souza, Jose H. Saito, Nelson D. A. Mascarenhas (In Memorian)

We propose a new Bayesian filter for speckle noise in Synthetic Aperture Radar (SAR) with one-look and quadratic detection. The non-local means filter will be applied over the noisy data modeled by that distribution, using stochastic distances, substituting the original Euclidean distance. We will use non-reference quality measures and we will compare the results with those obtained by state-of-the-art filters.

DOCUMENTO COMPLETO

• A Proposition of a Bayesian Filter for Despeckling SAR Images Using Conjugate Distribution, Stochastic Distances, and Non-Local Means