Pointwise Shape-Adaptive DCT for High-Quality Denoising and Deblocking of Grayscale and Color Images

By Alessandro Foi, Vladimir Katkovnik, Karen O. Egiazarian. IEEE Transactions on Image Processing, 2007

788 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition, Media Technology

View Paper: https://www.semanticscholar.org/paper/60e7d2cf0aa8088a9ba7a01b0aad19a61e441e0c

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

Real Image Denoising With Feature Attention (2019, neutral): “Due to their success, many variants were proposed, including SADCT [27], SAPCA [20], NLB [37], and INLM [29] which seek self-similar patches in different transform domains.”
A Modified Adaptive PCA Learning based Method for Image Denoising (2013, methodology): “[8] applied a shapeadaptive discrete cosine transform (DCT) to the neighborhood, which can achieve very sparse representation of the image and hence lead to effective denoising.”
Denoising of Hyperspectral Imagery Using Principal Component Analysis and Wavelet Shrinkage (2011, methodology): “This includes the Bayes least squares-Gaussian scale mixtures (BLS-GSM) method [19], the bivariate Laplacian mixture technique [20], and the local polynomial approximation-intersection of confidence intervals technique [21].”
Image Noise Level Estimation by Principal Component Analysis (2013, methodology): “We utilized the denoising method [2], which outperforms the methods [46]–[49] and can be considered as the state of the art in image denoising.”
Enhancing Quality for HEVC Compressed Videos (2017, methodology): “[4] applied pointwise Shape-Adaptive Discrete Cosine Transform (SA-DCT) to reduce blocking and ringing effects caused by JPEG compression.”
BM3D Image Denoising with Shape-Adaptive Principal Component Analysis (2009, methodology): “…regression [16] (also combined with learned dictionaries [2]); shape-adaptive DCT (SA-DCT) on neighborhoods whose shapes are adaptive to the image structures [7]; nonlocal (cid:222) ltering based on the assumption that there exist mutually similar blocks in natural images [1], [12], [3], [4].”
Image Denoising by Sparse 3-D Transform-Domain Collaborative Filtering (2007, methodology): “Another approach [9] is to exploit a shape-adaptive transform on neighborhoods whose shapes are adaptive to salient”
On learning optimized reaction diffusion processes for effective image restoration (2015, background): “This is a notable im-q JPEG decoder TGV [5] Dic-SR[6] SADCT[15] RTF[ provement as few methods can surpass BM3D more than 0.3dB in average [26].”
Sparsity-based image denoising via dictionary learning and structural clustering (2011, background): “The denoising results of three benchmark schemes (K-SVD [2], SA-DCT [ 18 ] and BM3D [5]) are all based on the source codes or executables released by the original authors.”
The SURE-LET Approach to Image Denoising (2007, methodology): “pointwise SA-DCT 8 [65], which is the application of their grayscale shape-adaptive DCT denoiser in the opponent color space, but using the adaptive neighborhoods defined in the luminance channel for all channels.”
Multi-frame Quality Enhancement for Compressed Video (2018, background): “Recently, there has been increasing interest in enhancing the visual quality of compressed image/video [27, 10, 18, 36, 6, 9, 38, 12, 24, 24, 42].”
Compression Artifacts Reduction by a Deep Convolutional Network (2015, methodology): “We ﬁrst compare AR-CNN with SA-DCT [5], which is widely regarded as the state-of-the-art deblocking oriented method [11, 14].”
SwinIR: Image Restoration Using Swin Transformer (2021, methodology): “Following [98, 88], we test different methods on two benchmark datasets (Classic5 [22] and LIVE1 [67]) for JPEG quality factors 10, 20, 30 and 40.”
Iterative Weighted Maximum Likelihood Denoising With Probabilistic Patch-Based Weights (2009, background): “They can be improved by using shape-adaptive domains [14],”
LIME: Low-Light Image Enhancement via Illumination Map Estimation (2017, background): “Hence, deblocking techniques [39], [40], [41], [42] may be required.”
CAS-CNN: A deep convolutional neural network for image compression artifact suppression (2016, methodology): “Testing of these networks was then performed on the LIVE1 dataset (29 images) [26] and, in case of AR-CNN, on the 5 test images of [15] and a self-collected dataset of 40 photographs from twitter as well.”
From Local Kernel to Nonlocal Multiple-Model Image Denoising (2009, methodology): “[37], the K-SVD algorithm by Aharon and Elad [3] and its multiscale extension (MSK-SVD) by Mairal et al.”