Déconvolution impulsionnelle positive myope

We consider in this paper the problem of the blind deconvolution of positive peaks, i.e. we wish to restore a sparse spike train whose peaks are positive and to estimate the impulse response and the problem hyperparameters. Because of the complexity of the posterior global criterion, we use an MCMC method based on the Gibbs sampler. Finally, we propose two new indexes to give an estimate of the signal, yielding an efficient analysis of the Markov chain.