Abstract:In this paper we integrate semi-local patches and the weighted graph Laplacian [17] into the framework of the low dimensional manifold model [12]. This approach is much faster than the original LDMM algorithm. The number of iterations is typically reduced from 100 to 10 and the equations in each step are much easier to solve. This new approach is tested in image inpainting and denoising and the results are better than the original LDMM and competitive with state-of-the-art methods.1. Introduction. In the past decade patch-based methods have achieved great success in image denoising, inpainting and other image processing problems [6, 2, 1, 4, 8, 22, 7, 24]. Among different patch-based nonlocal methods, the manifold model is attracting more and more attention, along with the development of manifold learning algorithms. The basic assumption in the manifold model is that the patches concentrate around a low dimensional smooth manifold. This assumption is verified in studies in image processing and computer vision [14, 15, 16, 9, 3, 13]. Recently the low dimensional manifold model (LDMM) was proposed for image processing [12], explicitly using the dimension of the patch manifold as a regularization to reconstruct the image. Using an elegant formula in differential geometry, LDMM is formulated as an optimization problem of minimizing the H1 norm of the coordinate functions of the patch manifold. The key step in the algorithm is to solve a Laplace-Beltrami equation on the patch set. This equation is solved by the point integral method [10, 11, 18, 19]. LDMM has been shown to have very good performance in image processing, especially …

Impulsive Noise and Gaussian-Impulsive Mixed Noise 1 Removal by Low Dimensional Manifold Model ∗

A Novel Image Classification Method Based on Manifold Learning and Gaussian Mixture Model

Low Dimensional Manifold Model for Image Processing.

Low dimensional manifold model with semi-local patches

Low-light Level Image De-Noising Algorithm Based on PCA

Noise attenuation in a low-dimensional manifold

An Improved Non-local Means Image De-noising Algorithm Using Mahalanobis Distance

Dataset Denoising Based on Manifold Assumption

Exploiting Image Local and Nonlocal Consistency for Mixed Gaussian-Impulse Noise Removal

Multimanifold Bistructured Low Rank Representation of hyperspectral images

Nonlocal Patches based Gaussian Mixture Model for Image Inpainting

Restoration of Images Corrupted by Impulse Noise and Mixed Gaussian Impulse Noise using Blind Inpainting

Image Denoising with Gaussian Mixture Model

Hyperspectral Image Mixed Noise Removal Based on Multidirectional Low-Rank Modeling and Spatial–Spectral Total Variation

Generalization of the Weighted Nonlocal Laplacian in Low Dimensional Manifold Model

Infimal convolution of data discrepancies for mixed noise removal

Manifold Learning for Image Denoising

3D Point Cloud Denoising Using Graph Laplacian Regularization of a Low Dimensional Manifold Model

Combined higher order non-convex total variation with overlapping group sparsity for impulse noise removal

Scalable low dimensional manifold model in the reconstruction of noisy and incomplete hyperspectral images

Design a Metric Robust to Complicated High Dimensional Noise for Efficient Manifold Denoising