Penalized maximum likelihood image restoration with positivity constraints:multiplicative algorithms

In this paper, we propose a general method to devise maximum likelihood penalized (regularized) algorithms with positivity constraints. Moreover, we explain how to obtain 'product forms' of these algorithms. The algorithmic method is based on Kuhn?Tucker first-order optimality conditions. Its application domain is not restricted to the cases considered in this paper, but it can be applied to any convex objective function with linear constraints. It is specially adapted to the case of objective functions with a bounded domain, which completely encloses the domain of the (linear) constraints. The Poisson noise case typical of this last situation and the Gaussian additive noise case are considered and they are associated with various forms of regularization functions, mainly quadratic and entropy terms. The algorithms are applied to the deconvolution of synthetic images blurred by a realistic point spread function similar to that of Hubble Space Telescope operating in the far-ultraviolet and corrupted by noise. The effect of the relaxation on the convergence speed of the algorithms is analysed. The particular behaviour of the algorithms corresponding to different forms of regularization functions is described. We show that the 'prior' image is a key point in the regularization and that the best results are obtained with Tikhonov regularization with a Laplacian operator. The analysis of the Poisson process and of a Gaussian additive noise leads to similar conclusions. We bring to the fore the close relationship between Tikhonov regularization using derivative operators, and regularization by a distance to a 'default image' introduced by Horne (Horne K 1985 Mon. Not. R. Astron. Soc. 213 129?41).

[1]  L. Armijo.  Minimization of functions having Lipschitz continuous first partial derivatives. , 1966 .

[2]  William H. Richardson,et al.  Bayesian-Based Iterative Method of Image Restoration , 1972 .

[3]  E. Polak.  Introduction to linear and nonlinear programming , 1973 .

[4]  L. Lucy.  An iterative technique for the rectification of observed distributions , 1974 .

[5]  D. Rubin,et al.  Maximum likelihood from incomplete data via the EM - algorithm plus discussions on the paper , 1977 .

[6]  M. Hayes,et al.  Convergence of iterative nonexpansive signal reconstruction algorithms , 1981 .

[7]  D. Youla,et al.  Image Restoration by the Method of Convex Projections: Part 1ߞTheory , 1982, IEEE Transactions on Medical Imaging.

[8]  L. Shepp,et al.  Maximum Likelihood Reconstruction for Emission Tomography , 1983, IEEE Transactions on Medical Imaging.

[9]  Gérard Thomas,et al.  A positive optimal deconvolution procedure , 1983, ICASSP.

[10]  Michael I. Miller,et al.  The Use of Sieves to Stabilize Images Produced with the EM Algorithm for Emission Tomography , 1985, IEEE Transactions on Nuclear Science.

[11]  L. Shepp,et al.  A Statistical Model for Positron Emission Tomography , 1985 .

[12]  Keith Horne,et al.  Images of accretion discs – I. The eclipse mapping method , 1985 .

[13]  D. Titterington.  General structure of regularization procedures in image reconstruction , 1985 .

[14]  M. Daube-Witherspoon,et al.  An Iterative Image Space Reconstruction Algorthm Suitable for Volume ECT , 1986, IEEE Transactions on Medical Imaging.

[15]  L. J. Thomas,et al.  Noise and Edge Artifacts in Maximum-Likelihood Reconstructions for Emission Tomography , 1987, IEEE Transactions on Medical Imaging.

[16]  P. P. B. Eggermont,et al.  Multiplicative iterative algorithms for convex programming , 1990 .

[17]  M. Bertero.  Linear Inverse and III-Posed Problems , 1989 .

[18]  Guy Demoment,et al.  Image reconstruction and restoration: overview of common estimation structures and problems , 1989, IEEE Trans. Acoust. Speech Signal Process..

[19]  A R De Pierro.  Nonlinear relaxation methods for solving symmetric linear complementarity problems , 1990 .

[20]  Michael I. Miller,et al.  Maximum a posteriori estimation for SPECT using regularization techniques on massively-parallel computers , 1991, Conference Record of the 1991 IEEE Nuclear Science Symposium and Medical Imaging Conference.

[21]  D. Snyder.  Modifications of the Lucy-Richardson Iteration for Restoring Hubble Space Telescope Imagery , 1991 .

[22]  I. Csiszár.  Why least squares and maximum entropy? An axiomatic approach to inference for linear inverse problems , 1991 .

[23]  Rémy Prost,et al.  Iterative constrained deconvulution , 1991, Signal Process..

[24]  Aggelos K. Katsaggelos,et al.  Maximum Likelihood Identification and Restoration of Images Using the Expectation-Maximization Algorithm , 1991 .

[25]  Joseph A. O'Sullivan,et al.  Deblurring subject to nonnegativity constraints , 1992, IEEE Trans. Signal Process..

[26]  A. R. De Pierro,et al.  On the relation between the ISRA and the EM algorithm for positron emission tomography , 1993, IEEE Trans. Medical Imaging.

[27]  P. L. Combettes.  The foundations of set theoretic estimation , 1993 .

[28]  Charles L. Byrne,et al.  Iterative image reconstruction algorithms based on cross-entropy minimization , 1993, IEEE Trans. Image Process..

[29]  Alfred O. Hero,et al.  Space-alternating generalized expectation-maximization algorithm , 1994, IEEE Trans. Signal Process..

[30]  Mario Bertero,et al.  Regularization methods in image restoration: An application to HST images , 1995, Int. J. Imaging Syst. Technol..

[31]  Alfred O. Hero,et al.  Ieee Transactions on Image Processing: to Appear Penalized Maximum-likelihood Image Reconstruction Using Space-alternating Generalized Em Algorithms , 2022 .

[32]  Alvaro R. De Pierro,et al.  On the convergence of an EM-type algorithm for penalized likelihood estimation in emission tomography , 1995, IEEE Trans. Medical Imaging.

[33]  Alvaro R. De Pierro,et al.  A modified expectation maximization algorithm for penalized likelihood estimation in emission tomography , 1995, IEEE Trans. Medical Imaging.

[34]  T. S. Zaccheo,et al.  Iterative maximum-likelihood estimators for positively constrained objects , 1996 .

[35]  Richard G. Lane.  Methods for maximum-likelihood deconvolution , 1996 .

[36]  Ken D. Sauer,et al.  A unified approach to statistical tomography using coordinate descent optimization , 1996, IEEE Trans. Image Process..

[37]  S. Twomey.  Introduction to the Mathematics of Inversion in Remote Sensing and Indirect Measurements , 1997 .

[38]  Mario Bertero,et al.  Introduction to Inverse Problems in Imaging , 1998 .

[39]  Claude Aime,et al.  Maximum-likelihood constrained regularized algorithms: an objective criterion for the determination of regularization parameters , 1999, Remote Sensing.

[40]  Gérard Thomas,et al.  A deconvolution technique using optimal Wiener filtering and regularization , 1999, Signal Process..

[41]  H. Lantéri,et al.  COMPARISON BETWEEN ISRA AND RLA ALGORITHMS. USE OF A WIENER FILTER BASED STOPPING CRITERION , 1999 .

[42]  R. Nowak,et al.  A statistical multiscale framework for Poisson inverse problems , 2000, IEEE Trans. Inf. Theory.

[43]  H Lantéri,et al.  A general method to devise maximum-likelihood signal restoration multiplicative algorithms with non-negativity constraints , 2001, Signal Process..

[44]  Arnold Neumaier,et al.  Introduction to Numerical Analysis , 2001 .

[45]  C. Aime.  Teaching astronomical speckle techniques , 2001 .

[46]  H. Lantéri,et al.  Ringing reduction in image restoration algorithms using a constraint on the inferior bound of the solution , 2002, Signal Process..