Edge-based blur kernel estimation using patch priors

By Libin Sun, Sunghyun Cho, Jue Wang, James Hays. 2013

496 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

Neural Blind Deconvolution Using Deep Priors (2019, methodology): “Analogous to the alternating minimization steps in traditional blind deconvolution [2,27, 29, 41, 56], the network parameters of G k and G x can also be optimized in an alternating manner.”
A Comparative Study for Single Image Blind Deblurring (2016, methodology): “However, existing datasets do not consider several crucial factors, e.g., scene depth variation , sensor saturation and nonlinear camera response Blurred input image Cho & Lee [2] Krishnan et al. [15] Whyte et al. [43] Sun et al. [38] Xu et al. [45] Zhong et al. [49] Pan et al. [29] Perrone et al. [”
$L_0$ -Regularized Intensity and Gradient Prior for Deblurring Text Images and Beyond (2017, methodology): “We evaluate the proposed algorithm on the benchmark dataset [19].”
A Neural Approach to Blind Motion Deblurring (2016, methodology): “We do not assume that the size of the kernel is known, and always estimate k[n] within a fixed-size support (51 × 51 as is standard for the benchmark [3]).”
Image Deblurring via Extreme Channels Prior (2017, methodology): “Extensive methods take advantage of sharp edges to estimate blur kernel [10, 19, 21].”
Learning to Deblur (2014, methodology): “We evaluate our method on the benchmark dataset of [27], which is an extended version of the standard test set of [26].”
Bringing a Blurry Frame Alive at High Frame-Rate With an Event Camera (2018, methodology): “Methods using gradient based regularizers, such as Gaussian scale mixture [7], l1\l2 norm [14], edge-based patch priors [34, 39] and l0norm regularizer [37], have been proposed.”
Blind Deblurring Using Internal Patch Recurrence (2014, methodology): “The statistics indicate that our algorithm and the algorithm of Sun et al.[18], which are the only patch-based methods, outperform all other approaches by a large gap.”
Real-World Blur Dataset for Learning and Benchmarking Deblurring Algorithms (2020, methodology): “[40] introduced a synthetic dataset generated from 80 natural images and the eight blur kernels.”
Human-Aware Motion Deblurring (2019, background): “Most representative datasets [23, 39] were constructed on the basis of a simplified camera-driven assumption that camera disturbance dominates the cause of blur [45].”
Blind Image Deblurring Using Dark Channel Prior (2016, methodology): “We examine our method on two natural image deblurring datasets [16, 29] and compare it to state-of-the-art natural image deblurring methods.”
From Motion Blur to Motion Flow: A Deep Learning Solution for Removing Heterogeneous Motion Blur (2016, methodology): “Various estimators have been proposed for more robust kernel estimation, such as edge-extraction-based maximum-a-posteriori (MAP) [6, 34], gradient activation based MAP [10], variational Bayesian methods [21, 22, 46], etc .”