A nonlocal denoising algorithm for manifold-valued images using second order statistics

DOI10.1137/16M1087114zbMATH Open1365.68453OpenAlexW2963562194MaRDI QIDQ5266380FDOQ5266380

Authors: Friederike Laus, Mila Nikolova, Johannes Persch, Gabriele Steidl 

Publication date: 2 June 2017

Published in: SIAM Journal on Imaging Sciences (Search for Journal in Brave)

Abstract: Nonlocal patch-based methods, in particular the Bayes' approach of Lebrun, Buades and Morel (2013), are considered as state-of-the-art methods for denoising (color) images corrupted by white Gaussian noise of moderate variance. This paper is the first attempt to generalize this technique to manifold-valued images. Such images, for example images with phase or directional entries or with values in the manifold of symmetric positive definite matrices, are frequently encountered in real-world applications. Generalizing the normal law to manifolds is not canonical and different attempts have been considered. Here we focus on a straightforward intrinsic model and discuss the relation to other approaches for specific manifolds. We reinterpret the Bayesian approach of Lebrun et al. (2013) in terms of minimum mean squared error estimation, which motivates our definition of a corresponding estimator on the manifold. With this estimator at hand we present a nonlocal patch-based method for the restoration of manifold-valued images. Various proof of concept examples demonstrate the potential of the proposed algorithm.

Full work available at URL: https://arxiv.org/abs/1607.08481

Recommendations

• Optimization methods for manifold-valued image processing
• Recent advances in denoising of manifold-valued images
• A nonlocal Bayesian image denoising algorithm
• A second order nonsmooth variational model for restoring manifold-valued images
• Below the surface of the non-local Bayesian image denoising method

zbMATH Keywords

normal distributionsecond-order statisticsmanifold-valued imagesnonlocal patch-based

Mathematics Subject Classification ID

Bayesian inference (62F15) Image analysis in multivariate analysis (62H35) Image processing (compression, reconstruction, etc.) in information and communication theory (94A08) Computing methodologies for image processing (68U10) White noise theory (60H40)

Cites Work

• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• A Review of Image Denoising Algorithms, with a New One
• A Riemannian framework for tensor computing
• A nonlocal Bayesian image denoising algorithm
• A nonlocal denoising algorithm for manifold-valued images using second order statistics
• A second order nonsmooth variational model for restoring manifold-valued images
• A second-order TV-type approach for inpainting and denoising higher dimensional combined cyclic and vector space data
• Convex analysis and optimization in Hadamard spaces
• Gaussian Distributions on Riemannian Symmetric Spaces: Statistical Learning With Structured Covariance Matrices
• Heat kernel and analysis on manifolds
• Intrinsic analysis of statistical estimation
• Intrinsic statistics on Riemannian manifolds: Basic tools for geometric measurements
• Invariant Bayesian estimation on manifolds
• Is Denoising Dead?
• Iterative Weighted Maximum Likelihood Denoising With Probabilistic Patch-Based Weights
• Kernel density estimation on spaces of Gaussian distributions and symmetric positive definite matrices
• Large sample theory of intrinsic and extrinsic sample means on manifolds. I
• ML characterization of the multivariate normal distribution
• Maximum likelihood characterization of distributions
• Multi-Scale Patch-Based Image Restoration
• Nonlinear total variation based noise removal algorithms
• Nonlocal Operators with Applications to Image Processing
• Numerical Methods forp-Harmonic Flows and Applications to Image Processing
• On the convergence of gradient descent for finding the Riemannian center of mass
• Patch-Based Near-Optimal Image Denoising
• Patch-Ordering-Based Wavelet Frame and Its Use in Inverse Problems
• Probability, Convexity, and Harmonic Maps with Small Image I: Uniqueness and Fine Existence
• Quadrature errors, discrepancies, and their relations to halftoning on the torus and the sphere
• Regularizing flows for constrained matrix-valued images
• Restoration of manifold-valued images by half-quadratic minimization
• Riemannian Gaussian Distributions on the Space of Symmetric Positive Definite Matrices
• Riemannian \(L^{p}\) center of mass: existence, uniqueness, and convexity
• Riemannian center of mass and mollifier smoothing
• SURE guided Gaussian mixture image denoising
• Second order differences of cyclic data and applications in variational denoising
• Secrets of image denoising cuisine
• Solving Inverse Problems With Piecewise Linear Estimators: From Gaussian Mixture Models to Structured Sparsity
• THE TRANSFORMATION OF POISSON, BINOMIAL AND NEGATIVE-BINOMIAL DATA
• The Circular Normal Distribution: Theory and Tables
• The normal distribution in some constrained sample spaces
• Total cyclic variation and generalizations
• Total generalized variation in diffusion tensor imaging
• Total variation regularization for manifold-valued data

Cited In (24)

• Recent advances in denoising of manifold-valued images
• Nonlocal myriad filters for Cauchy noise removal
• Regularization with metric double integrals of functions with values in a set of vectors
• Multivariate myriad filters based on parameter estimation of Student-\(t\) distributions
• Gaussian Patch Mixture Model Guided Low-Rank Covariance Matrix Minimization for Image Denoising
• A Graph Framework for Manifold-Valued Data
• Diffusion generated methods for denoising target-valued images
• Infimal Convolution Coupling of First and Second Order Differences on Manifold-Valued Images
• An inexact semismooth Newton method on Riemannian manifolds with application to duality-based total variation denoising
• Below the surface of the non-local Bayesian image denoising method
• Restoration guarantee of image inpainting via low rank patch matrix completion
• Lifting methods for manifold-valued variational problems
• Image Denoising: The Deep Learning Revolution and Beyond—A Survey Paper
• Compressive learning for patch-based image denoising
• Priors with coupled first and second order differences for manifold-valued image processing
• Variational restoration of nonflat image features: Models and algorithms
• Denoising of sphere- and SO(3)-valued data by relaxed Tikhonov regularization
• Optimization methods for manifold-valued image processing
• Diffusion tensor regularization with metric double integrals
• Fenchel duality theory and a primal-dual algorithm on Riemannian manifolds
• Learning from small data sets: patch-based regularizers in inverse problems for image reconstruction
• Convergence of the Time Discrete Metamorphosis Model on Hadamard Manifolds
• Video denoising via empirical Bayesian estimation of space-time patches
• A nonlocal denoising algorithm for manifold-valued images using second order statistics

Uses Software

• MTEX

This page was built for publication: A nonlocal denoising algorithm for manifold-valued images using second order statistics

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5266380)