Search

International audience; This work focuses on a parameter-free joint piecewise smooth image denoising and contour detection. Formulated as the minimization of a discrete Mumford-Shah functional and estimated via a theoretically grounded alternating minimization scheme, the bottleneck of such a variational approach lies in the need to fine-tune their hyperparameters, while not having access to ground truth data.To that aim, a Stein-like strategy providing optimal hyperparameters is designed, based on the minimization of an unbiased estimate of the quadratic risk.Efficient and automated minimization of the estimate of the risk crucially relies on an unbiased estimate of the gradient of the risk with respect to hyperparameters. Its practical implementation is performed using a forward differentiation of the alternating scheme minimizing the Mumford-Shah functional, requiring exact differentiation of the proximity operators involved. Intensive numerical experiments are performed on synthetic images with different geometry and noise levels, assessing the accuracy and the robustness of the proposed procedure.The resulting parameter-free piecewise-smooth estimation and contour detection procedure, not requiring prior image processing expertise nor annotated data, can then be applied to real-world images.

In electrical impedance tomography, we aim to solve the conductivity within a target body through electrical measurements made on the surface of the target. This inverse conductivity problem is severely ill-posed, especially in real applications with only partial boundary data available. Thus regularization has to be introduced. Conventionally regularization promoting smooth features is used, however, the Mumford--Shah regularizer familiar for image segmentation is more appropriate for targets consisting of several distinct objects or materials. It is, however, numerically challenging. We show theoretically through $\Gamma$-convergence that a modification of the Ambrosio--Tortorelli approximation of the Mumford--Shah regularizer is applicable to electrical impedance tomography, in particular the complete electrode model of boundary measurements. With numerical and experimental studies, we confirm that this functional works in practice and produces higher quality results than typical regularizations employed in electrical impedance tomography when the conductivity of the target consists of distinct smoothly-varying regions., Comment: 28 pages, 7 figures

We propose a $\Gamma$-convergent discrete approximation of the Mumford-Shah functional. The discrete functionals act on functions defined on stationary stochastic lattices and take into account general finite differences through a non-convex potential. In this setting the geometry of the lattice strongly influences the anisotropy of the limit functional. Thus we can use statistically isotropic lattices and stochastic homogenization techniques to approximate the vectorial Mumford-Shah functional in any dimension., Comment: 47 pages, reorganized version

We give an alternative proof of the regularity, up to the loose end, of minimizers, resp. critical points of the Mumford-Shah functional when they are sufficiently close to the cracktip, resp. they consist of a single arc terminating at an interior point., Comment: 27 pages. v3: corrected typos, added proof of (8.1), corrected acknowledgements. To appear in Journal de Math\'ematiques Pures et Appliqu\'ees. For Errata see https://www.math.ias.edu/delellis/sites/math.ias.edu.delellis/files/Errata-crackip.pdf

Let [math] be a bounded piecewise [math] open set with convex corners, and let ¶ MS ( u ) : = ∫ Ω | ∇ u | 2 d x + α ℋ 1 ( J u ) + β ∫ Ω | u − g | 2 d x ¶ be the Mumford–Shah functional on the space [math] , where [math] and [math] . We prove that the function [math] such that ¶ − Δ u + β u = β g in Ω , ∂ u ∕ ∂ ν = 0 on ∂ Ω ¶ is a local minimizer of [math] with respect to the [math] -topology. This is obtained as an application of interior and boundary monotonicity formulas for a weak notion of quasiminimizers of the Mumford–Shah energy. The local minimality result is then extended to more general free discontinuity problems taking into account also boundary conditions.

In this paper we introduce a new phase field approximation of the Mumford-Shah functional similar to the well-known one from Ambrosio and Tortorelli. However, in our setting the phase field is allowed to be a function of bounded variation, instead of an $H^1$-function. In the context of image segmentation, we also show how this new approximation can be used for numerical computations, which contains a total variation minimization of the phase field variable, as it appears in many problems of image processing. A comparison to the classical Ambrosio-Tortorelli approximation, where the phase field is an $H^1$-function, shows that the new model leads to sharper phase fields., 32 pages, 4 figures, 1 table, 39 references

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 (DTI)), 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 priori restrictions on the edge set and we do not have to discretize the data space. We apply our method to DTI as well as Q-ball imaging. Using the DTI model, we obtain a segmentation of the corpus callosum on real data.

Feature similarity measures are indispensable for joint image reconstruction in multi-modality medical imaging, which enable joint multi-modal image reconstruction (JmmIR) by communication of feature information from one modality to another, and vice versa. In this work, we establish an image similarity measure in terms of image edges from Tversky's theory of feature similarity in psychology. For joint bi-modal image reconstruction (JbmIR), it is found that this image similarity measure is an extended Mumford-Shah functional with a-priori edge information proposed previously from the perspective of regularization approach. This image similarity measure consists of Hausdorff measures of the common and different parts of image edges from both modalities. By construction, it posits that two images are more similar if they have more common edges and fewer unique/distinctive features, and will not force the nonexistent structures to be reconstructed when applied to JbmIR. With the Gamma-approximation of the JbmIR functional, an alternating minimization method is proposed for the JbmIR of diffuse optical tomography and x-ray computed tomography. The performance of the proposed method is evaluated by three numerical phantoms. It is found that the proposed method improves the reconstructed image quality by more than 10% compared to single modality image reconstruction (SmIR) in terms of the structural similarity index measure (SSIM), Comment: 29 pages, 11 figures

In this paper, we propose a novel retinal layer boundary model for segmentation of optical coherence tomography (OCT) images. The retinal layer boundary model consists of 9 open parametric contours representing the 9 retinal layers in OCT images. An intensity-based Mumford-Shah (MS) variational functional is first defined to evolve the retinal layer boundary model to segment the 9 layers simultaneously. By making use of the normals of open parametric contours, we construct equal sized adjacent narrowbands that are divided by each contour. Regional information in each narrowband can thus be integrated into the MS energy functional such that its optimisation is robust against different initialisations. A statistical prior is also imposed on the shape of the segmented parametric contours for the functional. As such, by minimising the MS energy functional the parametric contours can be driven towards the true boundaries of retinal layers, while the similarity of the contours with respect to training OCT shapes is preserved. Experimental results on real OCT images demonstrate that the method is accurate and robust to low quality OCT images with low contrast and high-level speckle noise, and it outperforms the recent geodesic distance based method for segmenting 9 layers of the retina in OCT images.

The Ambrosio-Tortorelli functional is a phase-field approximation of the Mumford-Shah functional that has been widely used for image segmentation. The approximation has the advantages of being easy to implement, maintaining the segmentation ability, and $\Gamma$-converging to the Mumford-Shah functional. However, it has been observed in actual computation that the segmentation ability of the Ambrosio-Tortorelli functional varies significantly with different values of the parameter and it even fails to $\Gamma$-converge to the original functional for some cases. In this paper we present an asymptotic analysis on the gradient flow equation of the Ambrosio-Tortorelli functional and show that the functional can have different segmentation behavior for small but finite values of the regularization parameter and eventually loses its segmentation ability as the parameter goes to zero when the input image is treated as a continuous function. This is consistent with the existing observation as well as the numerical examples presented in this work. A selection strategy for the regularization parameter and a scaling procedure for the solution are devised based on the analysis. Numerical results show that they lead to good segmentation of the Ambrosio-Tortorelli functional for real images., Comment: 22 pages

The Mumford–Shah model has been one of the most influential models in image segmentation and denoising. The optimization of the multiphase Mumford–Shah energy functional has been performed using level sets methods that optimize the Mumford–Shah energy by evolving the level sets via the gradient descent. These methods are very slow and prone to getting stuck in local optima due to the use of gradient descent. After the reformulation of the 2-phase Mumford–Shah functional on a graph, several groups investigated the hierarchical extension of the graph representation to multi class. The discrete hierarchical approaches are more effective than hierarchical (or direct) multiphase formulation using level sets. However, they provide approximate solutions and can diverge away from the optimal solution. In this paper, we present a discrete alternating optimization for the discretized Vese–Chan approximation of the piecewise constant multiphase Mumford–Shah functional that directly minimizes the multiphase functional without recursive bisection on the labels. Our approach handles the nonsubmodularity of the multiphase energy function and provides a global optimum if the image estimation data term is known apriori.

In this paper, we present a method for the numerical minimization of the Mumford-Shah functional that is based on the idea of topological asymptotic expansions. The basic idea is to cover the expected edge set with balls of radius \epsilon > 0 and use the number of balls, multiplied with 2\epsilon, as an estimate for the length of the edge set. We introduce a functional based on this idea and prove that it converges in the sense of \Gamma-limits to the Mumford-Shah functional. Moreover, we show that ideas from topological asymptotic analysis can be used for determining where to position the balls covering the edge set. The results of the proposed method are presented by means of two numerical examples and compared with the results of the classical approximation due to Ambrosio and Tortorelli., Comment: 28 pages, 2 figures

Hierarchies, such as the tree of shapes, are popular representations for image simplification and segmentation thanks to their multiscale structures. Selecting meaningful level lines (boundaries of shapes) yields to simplify image while preserving intact salient structures. Many image simplification and segmentation methods are driven by the optimization of an energy functional, for instance the celebrated Mumford-Shah functional. In this paper, we propose an efficient approach to hierarchical image simplification and segmentation based on the minimization of the piecewise-constant Mumford-Shah functional. This method conforms to the current trend that consists in producing hierarchical results rather than a unique partition. Contrary to classical approaches which compute optimal hierarchical segmentations from an input hierarchy of segmentations, we rely on the tree of shapes, a unique and well-defined representation equivalent to the image. Simply put, we compute for each level line of the image an attribute function that characterizes its persistence under the energy minimization. Then we stack the level lines from meaningless ones to salient ones through a saliency map based on extinction values defined on the tree-based shape space. Qualitative illustrations and quantitative evaluation on Weizmann segmentation evaluation database demonstrate the state-of-the-art performance of our method., Pattern Recognition Letters, Elsevier, 2016

This paper is the first part of an ongoing project aimed at providing a local minimality criterion, based on a second variation approach, for the triple point configurations of the Mumford-Shah functional.

The aim of this paper is to prove the existence of extension operators for SBV functions from periodically perforated domains. This result will be the fundamental tool to prove the compactness in a non coercive homogenization problem. Le but de cet article est de prouver l’existence d’op erateurs de prolongation pour des fonctions en SBV d enies sur des domaines p eriodiquement perfor es. Ce r esultat sera l’outil fondamental pour prouver la compacit e dans un probl eme non coercitif d’homog en eisation.

This paper presents a level-set based approach for the simultaneous reconstruction and segmentation of the activity as well as the density distribution from tomography data gathered by an integrated SPECT/CT scanner. Activity and density distributions are modeled as piecewise constant functions. The segmenting contours and the corresponding function values of both the activity and the density distribution are found as minimizers of a Mumford-Shah like functional over the set of admissible contours and -- for fixed contours -- over the spaces of piecewise constant density and activity distributions which may be discontinuous across their corresponding contours. For the latter step a Newton method is used to solve the nonlinear optimality system. Shape sensitivity calculus is used to find a descent direction for the cost functional with respect to the geometrical variables which leads to an update formula for the contours in the level-set framework. A heuristic approach for the insertion of new components for the activity as well as the density function is used. The method is tested for synthetic data with different noise levels.

Variational models have been studied for image segmentation application since the Mumford-Shah functional was introduced in the late 1980s. In this paper, we focus on multiphase segmentation with a new regularization term that yields an unsupervised segmentation model. We propose a functional that automatically chooses a favorable number of phases as it segments the image. The primary driving force of the segmentation is the intensity fitting term while a phase scale measure complements the regularization term. We propose a fast, yet simple, brute-force numerical algorithm and present experimental results showing the robustness and stability of the proposed model.

In this letter, we propose a shape-based active contours method for segmentation, based on a piecewise-constant approximation of the Mumford-Shah (M-S) functional. The Chan-Vese (C-V) formalism in a level set framework is used to formulate our method; however no sign distance function (SDF) is employed in the method proposed here. This method has the topology-free segmentation associated with the C-V algorithm and adds faster convergence, less memory requirement and fast re-initialization. These properties make the algorithm very attractive for video segmentation.

We propose a constrained version of Mumford and Shah's (1989) segmentation model with an information-theoretic point of view in order to devise a systematic procedure to segment brain magnetic resonance imaging (MRI) data for parametric -Map and -weighted images, in both 2-D and 3D settings. Incorporation of a tuning weight in particular adds a probabilistic flavor to our segmentation method, and makes the 3-tissue segmentation possible. Moreover, we proposed a novel method to jointly segment the -Map and calibrate RF Inhomogeneity (JSRIC). This method assumes theaveragevalue of white matter is the same across transverse slices in the central brain region, and JSRIC is able to rectify the flip angles to generate calibrated -Maps. In order to generate an accurate -Map, the determination of optimal flip-angles and the registration of flip-angle images are examined. Our JSRIC method is validated on two human subjects in the 2D -Map modality and our segmentation method is validated by two public databases, BrainWeb and IBSR, of -weighted modality in the 3D setting.

We deal with a variational approach to the inverse crack problem, that is the detection and reconstruction of cracks, and other defects, inside a conducting body by performing boundary measurements of current and voltage type. We formulate such an inverse problem in a free-discontinuity problems framework and propose a novel method for the numerical reconstruction of the cracks by the available boundary data. The proposed method is amenable to numerical computations and it is justified by a convergence analysis, as the error on the measurements goes to zero. We further notice that we use the Γ-convergence approximation of the Mumford–Shah functional due to Ambrosio and Tortorelli as the required regularization term.

The Mumford-Shah functional is a general and quite popular variational model for image segmentation. In particular, it provides the possibility to represent regions by smooth approximations. In this paper, we derive a statistical interpretation of the full (piecewise smooth) Mumford-Shah functional by relating it to recent works on local region statistics. Moreover, we show that this statistical interpretation comes along with several implications. Firstly, one can derive extended versions of the Mumford-Shah functional including more general distribution models. Secondly, it leads to faster implementations. Finally, thanks to the analytical expression of the smooth approximation via Gaussian convolution, the coordinate descent can be replaced by a true gradient descent.

A new necessary minimality condition for the Mumford-Shah functional is derived by means of second order variations. It is expressed in terms of a sign condition for a nonlocal quadratic form on $H^1_0(\Gamma)$, $\Gamma$ being a submanifold of the regular part of the discontinuity set of the critical point. Two equivalent formulations are provided: one in terms of the first eigenvalue of a suitable compact operator, the other involving a sort of nonlocal capacity of $\Gamma$. A sufficient condition for minimality is also deduced. Finally, an explicit example is discussed, where a complete characterization of the domains where the second variation is nonnegative can be given., Comment: 30 pages

In this paper we study the crack initiation in a hyper-elastic body governed by a Griffith-type energy. We prove that, during a load process through a time-dependent boundary datum of the type t → tg(x) and in the absence of strong singularities (e.g., this is the case of homogeneous isotropic materials) the crack initiation is brutal, that is, a big crack appears after a positive time ti > 0. Conversely, in the presence of a point x of strong singularity, a crack will depart from x at the initial time of loading and with zero velocity. We prove these facts for admissible cracks belonging to the large class of closed one-dimensional sets with a finite number of connected components. The main tool we employ to address the problem is a local minimality result for the functional \(\epsilon(\nu, \Gamma)\,:=\int_{\Omega} f(x,\nabla v)\,{\rm d}x+k{\mathcal{H}}^{1} (\Gamma),\) where \(\Omega \subseteq {\mathbb{R}}^{2}\) , k > 0 and f is a suitable Caratheodory function. We prove that if the uncracked configuration u of Ω relative to a boundary displacement ψ has at most uniformly weak singularities, then configurations (uΓ, Γ) with \({\mathcal{H}}^{1} (\Gamma)\) small enough are such that \(\epsilon(u,\emptyset) < \epsilon(u_{\Gamma},\Gamma)\) .

In [Progress Math. 233 (2005)], David suggested the existence of a new type of global minimizers for the Mumford-Shah functional in . The singular set of such a new minimizer belongs to a three parameters family of sets . We first derive necessary conditions satisfied by global minimizers of this family. Then we are led to study the first eigenvectors of the Laplace-Beltrami operator with Neumann boundary conditions on subdomains of with three reentrant corners. The necessary conditions are constraints on the eigenvalue and on the ratios between the three singular coefficients of the associated eigenvector. We use numerical methods (Singular Functions Method and Moussaoui's extraction formula) to compute the eigenvalues and the singular coefficients. We conclude that there is no for which the necessary conditions are satisfied and this shows that the hypothesis was wrong.

We consider a conducting body which presents some (unknown) perfectly insulating defects, such as cracks or cavities, for instance. We aim to reconstruct the defects by performing measurements of current and voltage type on a (known and accessible) part of the boundary of the conductor. A crucial step in this reconstruction is the determination of the electrostatic potential inside the conductor, by the electrostatic boundary measurements performed. Since the defects are unknown, we state such a determination problem as a free-discontinuity problem for the electrostatic potential in the framework of special functions of bounded variation. We provide a characterisation of the looked for electrostatic potential and we approximate it with the minimum points of a sequence of functionals, which take also in account the error in the measurements. These functionals are related to the so-called Mumford–Shah functional, which acts as a regularizing term and allows us to prove existence of minimizers and Γ -convergence properties.

We study the asymptotic limit of obstacle problems for Mumford–Shah type functionals with pgrowth in periodically perforated domains via the Γ -convergence of the associated free-discontinuity energies. In the limit a non-trivial penalization term related to the 1-capacity of the reference hole appears if and only if the size of the perforation scales like en/(n−1), e being its periodicity. We give two different formulations of the obstacle problem to include also perforations with Lebesgue measure zero.

We propose a new shape-based modeling technique for applications in imaging problems. Given a collection of shape priors (a shape dictionary), we define our problem as choosing the right dictionary elements and geometrically composing them through basic set operations to characterize desired regions in an image. This is a combinatorial problem solving which requires an exhaustive search among a large number of possibilities. We propose a convex relaxation to the problem to make it computationally tractable. We take some major steps towards the analysis of the proposed convex program and characterizing its minimizers. Applications vary from shape-based characterization, object tracking, optical character recognition, and shape recovery in occlusion, to other disciplines such as the geometric packing problem.

We present a preconditioning of a generalized forward-backward splitting algorithm for finding a zero of a sum of maximally monotone operators $\sum_{i=1}^{n} A_i + B$ with $B$ cocoercive, involving only the computation of $B$ and of the resolvent of each $A_i$ separately. This allows in particular to minimize functionals of the form $\sum_{i=1}^n g_i + f$ with $f$ smooth, using only the computation of the gradient of $f$ and of the proximity operator of each $g_i$ separately. By adapting the underlying metric, such preconditioning can serve two practical purposes: first, it might accelerate the convergence, or second, it might simplify the computation of the resolvent of $A_i$ for some $i$. In addition, in many cases of interest, our preconditioning strategy allows the economy of storage and computation concerning some auxiliary variables. In particular, we show how this approach can handle large-scale, nonsmooth, convex optimization problems structured on graphs, which arises in many image processing or learning applications, and that it compares favorably to alternatives in the literature., Comment: 35 pages, 5 figures

In this paper it is shown that any regular critical point of the Mumford–Shah functional, with positive definite second variation, is an isolated local minimizer with respect to competitors which are sufficiently close in the L 1 -topology. A global minimality result in small tubular neighborhoods of the discontinuity set is also established.

In this paper, we propose a new variational model to segment an object belonging to a given shape space using the active contour method, a geometric shape prior and the Mumford-Shah functional. The core of our model is an energy functional composed by three complementary terms. The first one is based on a shape model which constrains the active contour to get a shape of interest. The second term detects object boundaries from image gradients. And the third term drives globally the shape prior and the active contour towards a homogeneous intensity region. The segmentation of the object of interest is given by the minimum of our energy functional. This minimum is computed with the calculus of variations and the gradient descent method that provide a system of evolution equations solved with the well-known level set method. We also prove the existence of this minimum in the space of functions with bounded variation. Applications of the proposed model are presented on synthetic and medical images.