Silvia Bonettini

A scaled gradient projection method for constrained image deblurring

A class of scaled gradient projection methods for optimization problems with simple constraints is considered. These iterative algorithms can be useful in variational approaches to image deblurring that lead to minimized convex nonlinear functions subject to non-negativity constraints and, in some cases, to an additional flux conservation constraint. A special gradient projection method is introduced that exploits effective scaling strategies and steplength updating rules, appropriately designed for improving the convergence rate. We give convergence results for this scheme and we evaluate its effectiveness by means of an extensive computational study on the minimization problems arising from the maximum likelihood approach to image deblurring. Comparisons with the standard expectation maximization algorithm and with other iterative regularization schemes are also reported to show the computational gain provided by the proposed method.

On the Convergence of Primal–Dual Hybrid Gradient Algorithms for Total Variation Image Restoration

In this paper we establish the convergence of a general primal–dual method for nonsmooth convex optimization problems whose structure is typical in the imaging framework, as, for example, in the Total Variation image restoration problems. When the steplength parameters are a priori selected sequences, the convergence of the scheme is proved by showing that it can be considered as an ε-subgradient method on the primal formulation of the variational problem. Our scheme includes as special case the method recently proposed by Zhu and Chan for Total Variation image restoration from data degraded by Gaussian noise. Furthermore, the convergence hypotheses enable us to apply the same scheme also to other restoration problems, as the denoising and deblurring of images corrupted by Poisson noise, where the data fidelity function is defined as the generalized Kullback–Leibler divergence or the edge preserving removal of impulse noise. The numerical experience shows that the proposed scheme with a suitable choice of the steplength sequences performs well with respect to state-of-the-art methods, especially for Poisson denoising problems, and it exhibits fast initial and asymptotic convergence.

An alternating extragradient method for total variation-based image restoration from Poisson data

Variational models are a valid tool for edge-preserving image restoration from data affected by Poisson noise. This paper deals with total variation and hypersurface regularization in combination with the Kullbach Leibler divergence as a data fidelity function. We propose an iterative method, based on an alternating extragradient scheme, which is able to solve, in a numerically stable way, the primal–dual formulation of both total variation and hypersurface regularization problems. In this method, tailored for general smooth saddle-point problems, the stepsize parameter can be adaptively computed so that the convergence of the scheme is proved under mild assumptions. In the numerical experience, we focus the attention on the artificial smoothing parameter that makes different the total variation and hypersurface regularization. A set of experiments on image denoising and deblurring problems is performed in order to evaluate the influence of this smoothing parameter on the stability of the proposed method and on the features of the restored images.

A convergent blind deconvolution method for post-adaptive-optics astronomical imaging

In this paper, we propose a blind deconvolution method which applies to data perturbed by Poisson noise. The objective function is a generalized Kullback–Leibler (KL) divergence, depending on both the unknown object and unknown point spread function (PSF), without the addition of regularization terms; constrained minimization, with suitable convex constraints on both unknowns, is considered. The problem is non-convex and we propose to solve it by means of an inexact alternating minimization method, whose global convergence to stationary points of the objective function has been recently proved in a general setting. The method is iterative and each iteration, also called outer iteration, consists of alternating an update of the object and the PSF by means of a fixed number of iterations, also called inner iterations, of the scaled gradient projection (SGP) method. Therefore, the method is similar to other proposed methods based on the Richardson–Lucy (RL) algorithm, with SGP replacing RL. The use of SGP has two advantages: first, it allows one to prove global convergence of the blind method; secondly, it allows the introduction of different constraints on the object and the PSF. The specific constraint on the PSF, besides non-negativity and normalization, is an upper bound derived from the so-called Strehl ratio (SR), which is the ratio between the peak value of an aberrated versus a perfect wavefront. Therefore, a typical application, but not a unique one, is to the imaging of modern telescopes equipped with adaptive optics systems for the partial correction of the aberrations due to atmospheric turbulence. In the paper, we describe in detail the algorithm and we recall the results leading to its convergence. Moreover, we illustrate its effectiveness by means of numerical experiments whose results indicate that the method, pushed to convergence, is very promising in the reconstruction of non-dense stellar clusters. The case of more complex astronomical targets is also considered, but in this case regularization by early stopping of the outer iterations is required. However, the proposed method, based on SGP, allows generalization to the case of differentiable regularization terms added to the KL divergence, even if this generalization is outside the scope of this paper.

Nonnegative image reconstruction from sparse Fourier data: a new deconvolution algorithm

This paper deals with image restoration problems where the data are nonuniform samples of the Fourier transform of the unknown object. We study the inverse problem in both semidiscrete and fully discrete formulations, and our analysis leads to an optimization problem involving the minimization of the data discrepancy under nonnegativity constraints. In particular, we show that such a problem is equivalent to a deconvolution problem in the image space. We propose a practical algorithm, based on the gradient projection method, to compute a regularized solution in the discrete case. The key point in our deconvolution-based approach is that the fast Fourier transform can be employed in the algorithm implementation without the need of preprocessing the data. A numerical experimentation on simulated and real data from the NASA RHESSI mission is also performed.

Variable Metric Inexact Line-Search-Based Methods for Nonsmooth Optimization

We develop a new proximal-gradient method for minimizing the sum of a differentiable, possibly nonconvex, function plus a convex, possibly nondifferentiable, function. The key features of the proposed method are the definition of a suitable descent direction, based on the proximal operator associated to the convex part of the objective function, and an Armijo-like rule to determine the stepsize along this direction ensuring the sufficient decrease of the objective function. In this frame, we especially address the possibility of adopting a metric which may change at each iteration and an inexact computation of the proximal point defining the descent direction. For the more general nonconvex case, we prove that all limit points of the iterates sequence are stationary, while for convex objective functions we prove the convergence of the whole sequence to a minimizer, under the assumption that a minimizer exists. In the latter case, assuming also that the gradient of the smooth part of the objective function...

New convergence results for the scaled gradient projection method

The aim of this paper is to deepen the convergence analysis of the scaled gradient projection (SGP) method, proposed by Bonettini et al. in a recent paper for constrained smooth optimization. The main feature of SGP is the presence of a variable scaling matrix multiplying the gradient, which may change at each iteration. In the last few years, an extensive numerical experimentation showed that SGP equipped with a suitable choice of the scaling matrix is a very effective tool for solving large scale variational problems arising in image and signal processing. In spite of the very reliable numerical results observed, only a weak, though very general, convergence theorem is provided, establishing that any limit point of the sequence generated by SGP is stationary. Here, under the only assumption that the objective function is convex and that a solution exists, we prove that the sequence generated by SGP converges to a minimum point, if the scaling matrices sequence satisfies a simple and implementable condition. Moreover, assuming that the gradient of the objective function is Lipschitz continuous, we are also able to prove the O(1/k) convergence rate with respect to the objective function values. Finally, we present the results of a numerical experience on some relevant image restoration problems, showing that the proposed scaling matrix selection rule performs well also from the computational point of view.

Scaling techniques for gradient projection-type methods in astronomical image deblurring

The aim of this paper is to present a computational study on scaling techniques in gradient projection-type (GP-type) methods for deblurring of astronomical images corrupted by Poisson noise. In this case, the imaging problem is formulated as a non-negatively constrained minimization problem in which the objective function is the sum of a fit-to-data term, the Kullback–Leibler divergence, and a Tikhonov regularization term. The considered GP-type methods are formulated by a common iteration formula, where the scaling matrix and the step-length parameter characterize the different algorithms. Within this formulation, both first-order and Newton-like methods are analysed, with particular attention to those implementation features and behaviours relevant for the image restoration problem. The numerical experiments show that suited scaling strategies can enable the GP methods to quickly approximate accurate reconstructions and then are useful for designing effective image deblurring algorithms.

A New Semiblind Deconvolution Approach for Fourier-Based Image Restoration: An Application in Astronomy

The aim of this paper is to develop a new optimization algorithm for the restoration of an image starting from samples of its Fourier transform, when only partial information about the data frequencies is provided. The corresponding constrained optimization problem is approached with a cyclic block alternating scheme, in which projected gradient methods are used to find a regularized solution. Our algorithm is then applied to the imaging of high-energy radiation emitted during a solar flare through the analysis of the photon counts collected by the NASA Reuven Ramaty High Energy Solar Spectroscopic Imager satellite. Numerical experiments on simulated data show that, in both the presence and absence of statistical noise, the proposed approach provides some improvements in the reconstructions.

Some iterative methods for the solution of a symmetric indefinite KKT system

Abstract This paper is concerned with the numerical solution of a Karush–Kuhn–Tucker system. Such symmetric indefinite system arises when we solve a nonlinear programming problem by an Interior-Point (IP) approach. In this framework, we discuss the effectiveness of two inner iterative solvers: the method of multipliers and the preconditioned conjugate gradient method. We discuss the implementation details of these algorithms in an IP scheme and we report the results of a numerical comparison on a set of large scale test-problems arising from the discretization of elliptic control problems.