Abstract:We consider the restoration of piecewise constant images where the number of the regions and their values are not fixed in advance, with a good difference of piecewise constant values between neighboring regions, from noisy data obtained at the output of a linear operator (e.g., a blurring kernel or a Radon transform). Thus we also address the generic problem of unsupervised segmentation in the context of linear inverse problems. The segmentation and the restoration tasks are solved jointly by minimizing an objective function (an energy) composed of a quadratic data-fidelity term and a nonsmooth nonconvex regularization term. The pertinence of such an energy is ensured by the analytical properties of its minimizers. However, its practical interest used to be limited by the difficulty of the computational stage which requires a nonsmooth nonconvex minimization. Indeed, the existing methods are unsatisfactory since they (implicitly or explicitly) involve a smooth approximation of the regularization term and often get stuck in shallow local minima. The goal of this paper is to design a method that efficiently handles the nonsmooth nonconvex minimization. More precisely, we propose a continuation method where one tracks the minimizers along a sequence of approximate nonsmooth energies $\{J_\eps\}$, the first of which being strictly convex and the last one the original energy to minimize. Knowing the importance of the nonsmoothness of the regularization term for the segmentation task, each $J_\eps$ is nonsmooth and is expressed as the sum of an $\ell_1$ regularization term and a smooth nonconvex function. Furthermore, the local minimization of each $J_{\eps}$ is reformulated as the minimization of a smooth function subject to a set of linear constraints. The latter problem is solved by the modified primal-dual interior point method, which guarantees the descent direction at each step. Experimental results are presented and show the effectiveness and the efficiency of the proposed method. Comparison with simulated annealing methods further shows the advantage of our method.

Convergence Of An Iterative Nonlinear Scheme For Denoising Of Piecewise Constant Images

PIECEWISE CONSTANT SIGNAL AND IMAGE DENOISING USING A SELECTIVE AVERAGING METHOD WITH MULTIPLE NEIGHBORS

Image denoising via nonlinear hybrid diffusion

Nonlinear regularized reaction-diffusion filters for denoising of images with textures.

Efficient Reconstruction of Piecewise Constant Images Using Nonsmooth Nonconvex Minimization

IMAGE RESTORATION FROM NOISY INCOMPLETE FREQUENCY DATA BY ALTERNATIVE ITERATION SCHEME

Nonlinear Diffusion Driven by Local Features for Image Denoising

Low-cost denoising and deblurring using a novel nonlinear diffusion technique

Iterative adaptive weighted-mean filter for image denoising

Combined curvelet shrinkage and nonlinear anisotropic diffusion.

Curvelet Regularized Nonlinear Anisotropic Diffusion

An Improved Nonlinear Diffusion Algorithm for Image Denoising

Convergence Analysis for Iterative Data-Driven Tight Frame Construction Scheme

Image Denoising by Random Walk with Restart Kernel and Non-Subsampled Contourlet Transform

Iterative Approach to Image Compression with Noise : Optimizing Spatial and Tonal Data

Convergent Regularization in Inverse Problems and Linear Plug-and-Play Denoisers

A General Selective Averaging Method for Piecewise Constant Signal and Image Processing

A new development of non-local image denoising using fixed-point iteration for non-convex ℓp sparse optimization

A Globally Convergent Algorithm for a Constrained Non-Lipschitz Image Restoration Model

A New Feature-preserving Nonlinear Anisotropic Diffusion Method for Image Denoising.

An Alternating Iterative Algorithm for Image Deblurring and Denoising Problems