Imaging reconstruction problems are usually addressed by minimizing a noise-dependent data-misfit term plus an l_1 penalty term (or an l_1/l_2 penalty) which is chosen to enforce the sparsity (or the grouped-sparsity) of image coefficients on a given suitable basis. In this work we propose an adaptive regularization method for Poisson imaging which promotes (grouped) sparsity in a consistent manner and it overcomes the need for expensive optimization strategies commonly applied with Poisson measurements.

This presentation is part of Minisymposium “MS71 - Nonlinear and adaptive regularization for image restoration”
organized by: Claudio Estatico (University of Genoa) , Giuseppe Rodriguez (University of Cagliari) .

Authors:

Federico Benvenuto (University of Genoa)

Sabrina Guastavino (University of Genoa)

Keywords:

adaptive regularization, image deblurring, image reconstruction, image representation, inverse problems, machine learning, nonlinear optimization