A new bayesian Poisson denoising algorithm based on nonlocal means and stochastic distances

作者：

Highlights：

• A novel, mathematically formal and computationally efficient bayesian approach to Poisson denoising.

• It is based on the conjugacy of the Poisson and Gamma distributions, avoiding high cost procedures.

• The method is applied to denoising low-dose CT sinograms, using a non-local means filtering algorithm, replacing the euclidean distances by stochastic distances for the Gamma distribution.

• The proposed algorithm is competitive with several important algorithms of the literature.

摘要

•A novel, mathematically formal and computationally efficient bayesian approach to Poisson denoising.•It is based on the conjugacy of the Poisson and Gamma distributions, avoiding high cost procedures.•The method is applied to denoising low-dose CT sinograms, using a non-local means filtering algorithm, replacing the euclidean distances by stochastic distances for the Gamma distribution.•The proposed algorithm is competitive with several important algorithms of the literature.

论文关键词：Poisson denoising,Nonlocal means,Stochastic distances,Bayesian estimation,Conjugate distributions,Low dose CT

论文评审过程：Received 26 May 2020, Revised 30 September 2021, Accepted 3 October 2021, Available online 4 October 2021, Version of Record 10 October 2021.

论文官网地址：https://doi.org/10.1016/j.patcog.2021.108363