Détection-estimation de processus Bernoulli-gaussien-généralisés à la présence d'un bruit coloré non-gaussien

In this paper we address the problem of restoration of the wavelet coefficients related to crackle assumed to be a pulse respiratory signal for which a non-gaussian colored noise is added. This task, requiring multivariate probability density computations for the data likelihood term, often faces with the lack of analytical multidimensional expressions in the non-gaussian case. In this work, we propose a new approach based on copula theory to compute multivariate generalized Gaussian marginals to deal with the non-gaussianity of the wavelet coefficients of the colored added noise.