Statistical Multiresolution Dantzig Estimation in Imaging: Fundamental Concepts and Algorithmic Framework

In this paper we are concerned with fully automatic and locally adaptive estimation of functions in a "signal + noise"-model where the regression function may additionally be blurred by a linear operator, e.g. by a convolution. To this end, we introduce a general class of statistical multiresolution estimators and develop an algorithmic framework for computing those. By this we mean estimators that are defined as solutions of convex optimization problems with supremum-type constraints. We employ a combination of the alternating direction method of multipliers with Dykstra's algorithm for computing orthogonal projections onto intersections of convex sets and prove numerical convergence. The capability of the proposed method is illustrated by various examples from imaging and signal detection.

[1]  Markus Grasmair,et al.  The Equivalence of the Taut String Algorithm and BV-Regularization , 2006, Journal of Mathematical Imaging and Vision.

[2]  J. Polzehl,et al.  Adaptive weights smoothing with applications to image restoration , 1998 .

[3]  Douglas M. Hawkins,et al.  A Note on the Transformation of Chi-Squared Variables to Normality , 1986 .

[4]  R. Dykstra,et al.  A Method for Finding Projections onto the Intersection of Convex Sets in Hilbert Spaces , 1986 .

[5]  Axel Munk,et al.  Shape constrained regularisation by statistical multiresolution for inverse problems: asymptotic analysis. , 2010 .

[6]  Bernard Rougé,et al.  TV Based Image Restoration with Local Constraints , 2003, J. Sci. Comput..

[7]  I. Ekeland,et al.  Convex analysis and variational problems , 1976 .

[8]  S. Geer,et al.  Locally adaptive regression splines , 1997 .

[9]  J. Pawley,et al.  Handbook of Biological Confocal Microscopy , 1990, Springer US.

[10]  L. Duembgen,et al.  Multiscale inference about a density , 2007, 0706.3968.

[11]  C. Vogel.  Computational Methods for Inverse Problems , 1987 .

[12]  Lutz Dümbgen,et al.  Confidence Bands for Isotonic Median Curves Using Sign Tests , 2004 .

[13]  Benjamin Yakir,et al.  Tail probabilities for the null distribution of scanning statistics , 1998 .

[14]  P. Bickel,et al.  SIMULTANEOUS ANALYSIS OF LASSO AND DANTZIG SELECTOR , 2008, 0801.1095.

[15]  Yves Meyer,et al.  Oscillating Patterns in Image Processing and Nonlinear Evolution Equations: The Fifteenth Dean Jacqueline B. Lewis Memorial Lectures , 2001 .

[16]  Hein Hundal,et al.  The rate of convergence of dykstra's cyclic projections algorithm: The polyhedral case , 1994 .

[17]  D. Mergel,et al.  Residual-based localization and quantification of peaks in X-ray diffractograms , 2007 .

[18]  Faming Liang,et al.  Statistical and Computational Inverse Problems , 2006, Technometrics.

[19]  S. Osher,et al.  Fast TV Regularization for 2D Maximum Penalized Likelihood Estimation , 2011 .

[20]  Emmanuel J. Candès,et al.  Templates for convex cone problems with applications to sparse signal recovery , 2010, Math. Program. Comput..

[21]  P. Davies,et al.  Densities, spectral densities and modality , 2004, math/0410071.

[22]  D. Dobson,et al.  Convergence of an Iterative Method for Total Variation Denoising , 1997 .

[23]  Antonin Chambolle,et al.  Image Decomposition into a Bounded Variation Component and an Oscillating Component , 2005, Journal of Mathematical Imaging and Vision.

[24]  Axel Munk,et al.  SHAO'S THEOREM ON THE MAXIMUM OF STANDARDIZED RANDOM WALK INCREMENTS FOR MULTIDIMENSIONAL ARRAYS , 2009 .

[25]  Nicolai Bissantz,et al.  On difference‐based variance estimation in nonparametric regression when the covariate is high dimensional , 2005 .

[26]  P. Davies,et al.  Nonparametric Regression, Confidence Regions and Regularization , 2007, 0711.0690.

[27]  Gareth M. James,et al.  DASSO: connections between the Dantzig selector and lasso , 2009 .

[28]  Thoralf Mildenberger.  A geometric interpretation of the multiresolution criterion in nonparametric regression , 2008 .

[29]  Xu Shusheng,et al.  Estimation of the convergence rate of Dykstra’s cyclic projections algorithm in polyhedral case , 2000 .

[30]  R. Mathar,et al.  A cyclic projection algorithm via duality , 1989 .

[31]  J. Craggs.  Applied Mathematical Sciences , 1973 .

[32]  Eero P. Simoncelli,et al.  Image quality assessment: from error visibility to structural similarity , 2004, IEEE Transactions on Image Processing.

[33]  S. Hell,et al.  Breaking the diffraction resolution limit by stimulated emission: stimulated-emission-depletion fluorescence microscopy. , 1994, Optics letters.

[34]  许树声.  ESTIMATION OF THE CONVERGENCE RATE OF DYKSTRA’S CYCLIC PROJECTIONS ALGORITHM IN POLYHEDRAL CASE , 2000 .

[35]  J. Doob.  Stochastic processes , 1953 .

[36]  Michael A. Saunders,et al.  Atomic Decomposition by Basis Pursuit , 1998, SIAM J. Sci. Comput..

[37]  Yiqiu Dong,et al.  Automated Regularization Parameter Selection in Multi-Scale Total Variation Models for Image Restoration , 2011, Journal of Mathematical Imaging and Vision.

[38]  Zakhar Kabluchko,et al.  Extremes of the standardized Gaussian noise , 2010, 1007.0312.

[39]  S. Hell.  Far-Field Optical Nanoscopy , 2007, Science.

[40]  P. Davies,et al.  Local Extremes, Runs, Strings and Multiresolution , 2001 .

[41]  Axel Munk,et al.  Locally adaptive image denoising by a statistical multiresolution criterion , 2012, Comput. Stat. Data Anal..

[42]  V. Spokoiny,et al.  Multiscale testing of qualitative hypotheses , 2001 .

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

[44]  Justin K. Romberg,et al.  The Dantzig selector and generalized thresholding , 2008, 2008 42nd Annual Conference on Information Sciences and Systems.

[45]  R. Tibshirani.  Regression Shrinkage and Selection via the Lasso , 1996 .

[46]  O. Scherzer,et al.  Regularization of ill-posed linear equations by the non- stationary augmented Lagrangian method , 2010 .

[47]  Terence Tao,et al.  The Dantzig selector: Statistical estimation when P is much larger than n , 2005, math/0506081.

[48]  Yong Zhang,et al.  An alternating direction method for finding Dantzig selectors , 2010, Comput. Stat. Data Anal..

[49]  V. Liebscher,et al.  Consistencies and rates of convergence of jump-penalized least squares estimators , 2009, 0902.4838.