Image Super-Resolution by TV-Regularization and Bregman Iteration

By Antonio Marquina, Stanley J. Osher. Journal of Scientific Computing, 2008

473 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/75224c681d1595865f179c15cd94911ed47b8274

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

Citations in the corpus

Sparse Representation Based Image Interpolation With Nonlocal Autoregressive Modeling (2013, background): “widely used regularizer is the total variation (TV) model [10-13], which assumes that natural images have”
Modern regularization methods for inverse problems (2018, background): “ularization methods have found applications in all kind of inverse problems and pushed forward the state of the art. As some examples let us mention TV/TGV Bregman iterations for superresolution (cf. [276]), PET reconstruction (cf. [285, 284]) or STED microscopy (cf. [62, 64]), as well as TGV recons”
MADNet: A Fast and Lightweight Network for Single-Image Super Resolution (2020, background): “Given the perception that MAE or MSE tends to lead a smooth result, we additionally introduce a total variation (TV) regularizer [10], [29] to constrain the smoothness of where ∇ h ( · ) and ∇ v ( · ) denote the gradient operator among the horizontal and vertical direction, respectively.”
Super-resolution: a comprehensive survey (2014, background): “The Tikhonov regularization term does not allow discontinuities in the solution and therefore may not recover the edges properly [354].”
Nonlocally Centralized Sparse Representation for Image Restoration (2013, methodology): “From left to right and top to bottom: original image, LR image, the reconstructed images by TV [26] (PSNR=31.”
Deep Learning for Single Image Super-Resolution: A Brief Review (2018, methodology): “Reconstruction-based SR methods [12], [13], [14], [15] often adopt sophisticated prior knowledge to restrict the possible solution space with an advantage of generating ﬂexible and sharp details.”
Denoising Prior Driven Deep Neural Network for Image Restoration (2018, background): “For example, the IR problem is a denoising problem [1]–[5] when A is an identical matrix and becomes a deblurring problem [6]–[9] when A is a blurring matrix/operator, or a super-resolution problem [8], [10]–[12] when A is a subsampling matrix/operator.”
Centralized sparse representation for image restoration (2011, methodology): “, the softcut method of [12], the TV-based method of [22] and the sparse representation based method of [27].”
Image Recovery via Nonlocal Operators (2010, unknown): “Due to its virtue of preserving edges, it is widely used in the many applications of image processing, such as blind deconvolution [7, 15, 20], inpainting [6] and superresolution [21].”
Image Deblurring and Super-Resolution by Adaptive Sparse Domain Selection and Adaptive Regularization (2010, methodology): “The average PSNR and SSIM values by the test methods are listed in Table VIII, and the distributions of PSNR gain of our method over [25] and [47] are shown in Fig.”
Image Super-Resolution With Sparse Neighbor Embedding (2012, methodology): “(h) Result of the TV-based method.”