PSF estimation using sharp edge prediction

By Neel Joshi, Richard Szeliski, David J. Kriegman. 2008 IEEE Conference on Computer Vision and Pattern Recognition, 2008

666 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition, Media Technology

View Paper: https://www.semanticscholar.org/paper/521807ebfa5b83afda0fe2ceced91f449d1464d8

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

Accurate Blur Models vs. Image Priors in Single Image Super-resolution (2013, background): “A few single-image SR works which do attempt to estimate or take the unknown kernel into account include [3, 24, 17, 10, 11, 12].”
Image deblurring and denoising using color priors (2009, background): “Result (b) is sharp but noisy and has ringing artifacts; (c) is overly-smoothed; our result (d) has the sharpness of Lucy-Richardson (most notably in the red/white checkerboard) with significantly reduced noise and ringing. ural images, has also been studied [8] [11].”
Understanding and evaluating blind deconvolution algorithms (2009, methodology): “Recent algorithms have proposed to address the ill-posedness of blind deconvolution by characterizing x using natural image statistics [16, 4, 14, 6, 7, 3, 20].”
Learning a convolutional neural network for non-uniform motion blur removal (2015, background): “Image deblurring [2, 6, 9, 11, 14, 19, 23, 30] aims at recovering sharp image from a blurry image due to camera shake, object motion or out-of-focus.”
Ieee Transactions on Pattern Analysis and Machine Intelligence 1 on Bayesian Adaptive Video Super Resolution (methodology): “Their analysis assumes no spatial prior on the blur kernel, while Joshi et al.[14] used a smoothness prior for the blur kernel and obtained reliable estimates.”
Blind Super-Resolution Kernel Estimation using an Internal-GAN (2019, methodology): “Some Blind-SR methods [33, 17, 14, 13, 3] were introduced prior to the deep learning era.”
A no-reference metric for evaluating the quality of motion deblurring (2013, methodology): “Several recent approaches [Joshi et al. 2008; Cho and Lee 2009; Xu and Jia 2010; Cho et al. 2011] rely on extracting edge profiles for PSF estimation, under the assumption that edges are sharp in natural images.”
On Bayesian Adaptive Video Super Resolution (2014, background): “Their analysis assumes no spatial prior on the blur kernel, while Joshi et al. [14] used a smoothness prior for the blur kernel and obtained reliable estimates.”
Blind image super-resolution with spatially variant degradations (2019, methodology): “…example, Qiao et al. [2006] propose an SVM to estimate the variance of a Gaussian blur kernel using features extracted with a Sobel operator, while Joshi et al. [2008] assume that contours in the image correspond to sharp edges that can be reconstructed, and the camera PSF is computed from these…”
Exploring photobios (2011, methodology): “1We note that the Gaussian is an imperfect PSF model [Joshi et al. 2008], but still useful as a first-order approximation.”
Analyzing spatially-varying blur (2010, background): “In the latter case, one can now reliably estimate and reverse the effects of blur through a variety of methods, even when the (spatially-uniform) kernel is of a complex, nonparametric form [4, 11, 12, 14, 17, 20].”
A face antispoofing database with diverse attacks (2012, background): “It is known that the printing, imaging and displaying process can introduce distortions such as blur and aliasing [2, 18].”
Nonparametric Blind Super-resolution (2013, neutral): “[12] both for deblurring and for SR.”
Blind Deblurring Using Internal Patch Recurrence (2014, background): “These range from setting a threshold on the image gradients [9] to shock and bilateral filtering [1,19,2].”
Learning to Estimate and Remove Non-uniform Image Blur (2013, background): “Chakrabarti et al. [5] show interesting results for separating the blur and sharp regions in an image, but do not address deblurring itself, which is equally challenging.”
Image Deblurring via Extreme Channels Prior (2017, methodology): “Extensive methods take advantage of sharp edges to estimate blur kernel [10, 19, 21].”
$L_0$ -Regularized Intensity and Gradient Prior for Deblurring Text Images and Beyond (2017, methodology): “In addition, methods that explicitly select sharp edges for kernel estimation have been proposed [3], [4], [17] with demonstrated success on benchmark datasets [18].”
Blind Super-Resolution With Iterative Kernel Correction (2019, methodology): “Most existing blind SR methods are model-based [3, 32, 11, 12, 14], which usually involve complicated optimization procedures.”
Unnatural L0 Sparse Representation for Natural Image Deblurring (2013, methodology): “[11, 13, 19, 3, 25, 9, 23], which are referred to as semi-blind schemes, and those [20, 16] implicitly incorporating special regularization to remove detrimental structures in early stages and gradually enrich image details in iterations.”