A Review of Image Denoising Algorithms, with a New One

By A. Buades, Bartomeu Coll, Jean-Michel Morel. 2005

4,615 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition, Media Technology

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

In Media Technology

Citations in the corpus

Adversarial Autoencoders (2015, methodology): “Several methods exist for quantitatively analyzing the quality of a generative model (Buades et al., 2005; Lyu & Simoncelli, 2009).”
The emerging field of signal processing on graphs: Extending high-dimensional data analysis to networks and other irregular domains (2012, neutral): “…The basic graph spectral ﬁltering (12) can be used to implement discrete versions of well-known continuous ﬁltering techniques such as Gaussian smoothing, bilateral ﬁltering, total variation ﬁltering, anisotropic diffusion, and non-local means ﬁltering (see, e.g., [39] and references therein).”
Nonlocally Centralized Sparse Representation for Image Restoration (2013, methodology): “Based on the fact that natural images often contain repetitive structures, i.e., the rich amount of nonlocal redundancies [30], we search the nonlocal similar patches to the given patch i in a large window centered at pixel i .”
Multi-scale dictionary for single image super-resolution (2012, methodology): “For the non-local similarity regularization term, we rewrite it in the following formˆX = arg min{ i∈X X i − c N i S(x i ) 2 2 }, (17) where S(x i ) represents a column vector by stacking similar pixels centered at x i in lexicographical ordering and c N i is the row vector consisting of the corresp”
Robust video denoising using low rank matrix completion (2010, methodology): “The proposed video denoising method is built upon the same methodology “grouping and collaboratively filtering” as many patch-based methods (e.g. [2, 4, 10, 11]) do.”
Image denoising with nonlocal spectral graph wavelets (2010, neutral): “[5].”
Super-resolution from a single image (2009, methodology): “Our approach attempts to recover at each pixel its best possible resolution increase based on its patch redundancy within and across scales.”
Denoising of 3 D Magnetic Resonance Images by Using 1 Higher-Order Singular Value Decomposition 2 (2014, methodology): “A simple and classic method is the 13 well-known nonlocal means (NLM) filter (Buades et al., 2005), which averages 14 multiple pixels in a large search window and avoids the smoothing of edges by 15 assigning only high weights to pixels with similar local patterns.”
MRI denoising using Non-Local Means (2008, methodology): “As there is no objective criteria fully meeting such requirements, visual inspection of the image residuals (i.e. the di ﬀ erence between the original and the ﬁltered image) was also used to evaluate ﬁlter e ﬃ cacy: the so called method noise proposed by Buades et al. (2005).”
Sparse Representation for Computer Vision and Pattern Recognition (2010, background): “State-of-the-art results obtained in [51] are “shared” with those in [19], which extends the non-local means approach developed in [5], [12].”
Nonlocal Discrete Regularization on Weighted Graphs: A Framework for Image and Manifold Processing (2008, methodology): “In this case, one iteration of (25) corresponds to the nonlocal means (NLMeans) filter [1] (see also [23] and [24]).”
Learning Multiscale Sparse Representations for Image and Video Restoration (2008, methodology): “The proposed algorithm also competes favorably with the most recent and state-of-the-art result in this field [12], which is based on the nonlocal means algorithm [4].”
Video denoising by sparse 3D transform-domain collaborative filtering (2007, methodology): “A recent denoising strategy, the non-local spatial estimation [6], has also been adapted to video denoising [7].”
Noise2Self: Blind Denoising by Self-Supervision (2019, background): “The classic non-local means algo-rithm replaces the center pixel of each patch with a weighted average of central pixels from similar patches (Buades et al., 2005a).”
Patch-Based Nonlocal Functional for Denoising Fluorescence Microscopy Image Sequences (2010, methodology): “To overcome this problem, sophisticated methods (see [1]) but designed for still images have been adapted to videos.”