Mumford-Shah and Potts regularization for manifold-valued data

DOI10.1007/S10851-015-0628-2zbMATH Open1344.49055arXiv1410.1699OpenAlexW3125749597MaRDI QIDQ294412FDOQ294412

Authors: Andreas Weinmann, Laurent Demaret, Martin Storath 

Publication date: 16 June 2016

Published in: Journal of Mathematical Imaging and Vision (Search for Journal in Brave)

Abstract: Mumford-Shah and Potts functionals are powerful variational models for regularization which are widely used in signal and image processing; typical applications are edge-preserving denoising and segmentation. Being both non-smooth and non-convex, they are computationally challenging even for scalar data. For manifold-valued data, the problem becomes even more involved since typical features of vector spaces are not available. In this paper, we propose algorithms for Mumford-Shah and for Potts regularization of manifold-valued signals and images. For the univariate problems, we derive solvers based on dynamic programming combined with (convex) optimization techniques for manifold-valued data. For the class of Cartan-Hadamard manifolds (which includes the data space in diffusion tensor imaging), we show that our algorithms compute global minimizers for any starting point. For the multivariate Mumford-Shah and Potts problems (for image regularization) we propose a splitting into suitable subproblems which we can solve exactly using the techniques developed for the corresponding univariate problems. Our method does not require any a priori restrictions on the edge set and we do not have to discretize the data space. We apply our method to diffusion tensor imaging (DTI) as well as Q-ball imaging. Using the DTI model, we obtain a segmentation of the corpus callosum.

Full work available at URL: https://arxiv.org/abs/1410.1699

Recommendations

• Non-smooth variational regularization for processing manifold-valued data
• Total variation regularization for manifold-valued data
• An algorithmic framework for Mumford-Shah regularization of inverse problems in imaging
• Total generalized variation for manifold-valued data
• An algorithm for second order Mumford-Shah models based on a Taylor jet formulation

zbMATH Keywords

regularizationconvex optimizationdynamic programmingdiffusion tensor imagingMumford-Shah functionalproximal methodsHadamard manifoldjump sparsityPotts functionalQ-ball imaging

Mathematics Subject Classification ID

Convex programming (90C25) Dynamic programming (90C39) Numerical methods based on nonlinear programming (49M37) Image processing (compression, reconstruction, etc.) in information and communication theory (94A08) Computing methodologies for image processing (68U10) Dynamic programming in optimal control and differential games (49L20)

Cites Work

• Title not available (Why is that?)
• Intrinsic statistics on Riemannian manifolds: Basic tools for geometric measurements
• Consistencies and rates of convergence of jump-penalized least squares estimators
• Stochastic Relaxation, Gibbs Distributions, and the Bayesian Restoration of Images
• Optimal detection of changepoints with a linear computational cost
• Numerical Methods forp-Harmonic Flows and Applications to Image Processing
• Geodesic regression and the theory of least squares on Riemannian manifolds
• Title not available (Why is that?)
• Large sample theory of intrinsic and extrinsic sample means on manifolds. I
• Title not available (Why is that?)
• Jump-Sparse and Sparse Recovery Using Potts Functionals
• Optimal approximations by piecewise smooth functions and associated variational problems
• Approximation of functional depending on jumps by elliptic functional via t-convergence
• Finite-differences discretizations of the mumford-shah functional
• Image Segmentation by Variational Methods: Mumford and Shah Functional and the Discrete Approximations
• Riemannian center of mass and mollifier smoothing
• A Riemannian framework for tensor computing
• On the convergence of gradient descent for finding the Riemannian center of mass
• Curve evolution implementation of the Mumford-Shah functional for image segmentation, denoising, interpolation, and magnification
• Title not available (Why is that?)
• Optimal a priori discretization error bounds for geodesic finite elements
• Riemannian geometry for the statistical analysis of diffusion tensor data
• Manifolds of nonpositive curvature
• Rectifiable Metric Spaces: Local Structure and Regularity of the Hausdorff Measure
• Iterative thresholding meets free-discontinuity problems
• Intrinsic analysis of statistical estimation
• Computing medians and means in Hadamard spaces
• Multiplier methods: A survey
• Existence of minimizers of the Mumford-Shah functional with singular operators and unbounded data
• Large sample theory of intrinsic and extrinsic sample means on manifolds. II.
• Convergence and \(C^1\) analysis of subdivision schemes on manifolds by proximity
• Exact algorithms for \(L^1\)-TV regularization of real-valued or circle-valued signals
• Interpolatory multiscale representation for functions between manifolds
• An algorithmic framework for Mumford-Shah regularization of inverse problems in imaging
• Probability, Convexity, and Harmonic Maps with Small Image I: Uniqueness and Fine Existence
• Title not available (Why is that?)
• A splitting method for orthogonality constrained problems
• Smoothers for Discontinuous Signals
• Fast partitioning of vector-valued images
• Total variation regularization for manifold-valued data
• Second order differences of cyclic data and applications in variational denoising
• Joint image reconstruction and segmentation using the Potts model
• The \(L^1\)-Potts functional for robust jump-sparse reconstruction
• Group-valued regularization for motion segmentation of articulated shapes
• On the Complexity of Mumford–Shah-Type Regularization, Viewed as a Relaxed Sparsity Constraint
• Complexity penalized least squares estimators: Analytical results
• Regularizing properties of the Mumford–Shah functional for imaging applications
• Multiscale Representations for Manifold-Valued Data
• On approximating the Riemannian 1-center
• Interpolatory wavelets for manifold-valued data

Cited In (15)

• Mumford-Shah regularization in electrical impedance tomography with complete electrode model
• An inexact semismooth Newton method on Riemannian manifolds with application to duality-based total variation denoising
• Jump-penalized least absolute values estimation of scalar or circle-valued signals
• Total generalized variation for manifold-valued data
• Wavelet Sparse Regularization for Manifold-Valued Data
• Lifting methods for manifold-valued variational problems
• Smoothing for signals with discontinuities using higher order Mumford-Shah models
• Non-smooth variational regularization for processing manifold-valued data
• Total variation regularization for manifold-valued data
• Measure-valued variational models with applications to diffusion-weighted imaging
• Manifold-valued data in medical imaging applications
• An algorithm for second order Mumford-Shah models based on a Taylor jet formulation
• Lifting the convex conjugate in Lagrangian relaxations: a tractable approach for continuous Markov random fields
• Exact algorithms for \(L^1\)-TV regularization of real-valued or circle-valued signals
• First order methods for optimization on Riemannian manifolds

Uses Software

• pottslab
• Camino

This page was built for publication: Mumford-Shah and Potts regularization for manifold-valued data

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q294412)