Total variation denoising has attracted considerable attention in 1-D and 2-D signal processing. For image denoising, the convex cost function can be viewed as the regularized linear least squares problem (ℓ1 regularizer for anisotropic case and ℓ2 regularizer for isotropic case). However, these convex regularizers often underestimates the high-amplitude components of the true image. In this paper, non-convex regularizers for 2-D TV denoising models are proposed. These regularizers are based on the Moreau envelop and minimax-concave penalty which can maintain the convexity of the cost functions. Then efficient algorithms based on forward-backward splitting are proposed to solve the new cost functions. Numerical results show the effectiveness of the proposed non-convex regularizers for both synthetic and real-world image.
Researchers utilised various types of information in active contour models to define new energy functionals for image segmentation. These models aim to extract all potential objects from the background, but non‐target objects and noise are also obtained. In this study, the authors aim to extract target objects with sparse representation method. The original indicator function (a binary function) with respect to the level set function is used to represent the foreground (value is 1) and background(value is 0). From another point of view, an indicator function can be represented by linear combination of a set of the basis function. Firstly, by a label operator for the indicator function in each iteration, every connected area is represented by a basis function. Secondly, the linear combination of these basis functions is used to represent objects. Finally, through the sparsity constraint of coefficients of basis functions, the object extraction is viewed as a sparse representation problem. Meanwhile, a corresponding improved orthogonal matching pursuit algorithm is designed to obtain the ideal results. Experiments demonstrate that the proposed method has superior performance in object extraction compared with state‐of‐the‐art active contour models. Furthermore, the proposed method also increases the flexibility of applications.
20 July 2020 in International Journal of Imaging Systems and Technology
Sparse reconstruction has attracted considerable attention in recent years and shown powerful capabilities in many applications. In standard sparse reconstruction, the sparse nonzero elements appear anywhere in a vector. However, in many applications, the nonzero elements usually exhibit additional structure. Structured sparsity can be reconstructed from asymptotically less measurement than standard sparsity. In this paper, a unified framework is given to express the existing sparsity structures. Then efficient algorithm based on split Bregman iteration is proposed to solve the structured sparse reconstruction problem. The convergence of the proposed algorithm is also discussed. Numerical results show the effectiveness of the proposed algorithm for both synthetic and real-world data.
Traditional ℓ1-regularized compressed sensing magnetic resonance imaging (CS-MRI) model tends to underestimate the fine textures and edges of the MR image, which play important roles in clinical diagnosis. In contrast, the convex nonconvex (CNC) strategy allows the use of nonconvex regularization while maintaining the convexity of the total objective function. Plug-and-play (PnP) algorithm is a powerful framework for sparse regularization problems, which plug any advanced denoiser into traditional proximal algorithms. In this paper, we propose a PnP-ADMM algorithm for CS-MRI reconstruction with CNC sparse regularization. We first obtain the proximal operator for CNC sparse regularization. Then we present PnP-ADMM algorithm by replacing the proximal operator of ADMM with properly pre-trained denoisers. Furthermore, we conduct comparison experiments using various denoisers under different sampling templates for different images. The experimental results verify the effectiveness of the proposed algorithm with both numerical criteria and visual effects.
By allowing multiple L indices to be chosen per iteration, multiple orthogonal least squares (MOLS) have been proposed that is an extension of the classical greedy algorithm OLS. Wang and Li 2017 demonstrated that MOLS can successfully reconstruct K ‐sparse signals from compressed measurements in at most K iterations if the sensing matrix has unit ‐norm columns satisfying the restricted isometry property (RIP) of order LK with . In this study, by increasing the RIP order just by one (i.e. from LK ), the authors refine the bound further to . In the noisy case, they also propose a stopping criterion of MOLS algorithm, and with this criterion MOLS algorithm can recover the support of sparse signal successfully.