Blind Motion Deblurring Using Image Statistics

By Anat Levin. 2006

465 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

PSF estimation using sharp edge prediction (2008, background): “For more complex blurs, such as motion blur, many recent single-image estimation techniques model blurs as a collections of 1D or 2D box blurs and use segmentation techniques to handle multiple motions [7, 4, 2].”
Single Image Deblurring Using Motion Density Functions (2010, methodology): “[12] take a similar approach for object motion blur, where an image is segmented into several areas of different motion blur and then each area is deblurred independently.”
Photo and Video Quality Evaluation: Focusing on the Subject (2008, background): “[13] presented a scheme to identify blur in an image when the blur is caused by 1D motion.”
From Motion Blur to Motion Flow: A Deep Learning Solution for Removing Heterogeneous Motion Blur (2016, methodology): “To handle blur caused by object motion, some methods [20, 8, 15, 26] segment images into areas with different types of blur, and are thus heavily dependent on an accruate segmentation of a blurred image.”
Blind deconvolution in scientific imaging & computational photography (2011, methodology): “However, noise in the observed image renders deconvolution even an ill-posed problem, for perfect knowledge of the PSF, leading to unwanted noise amplification and ringing artifacts in the reconstructed image (Levin, 2006).”
Learning a convolutional neural network for non-uniform motion blur removal (2015, background): “[17] predicts motion blur kernel using natural image statistics.”
Image deblurring with blurred/noisy image pairs (2007, background): “In [Levin 2006], the image is segmented into several layers with different kernels.”
Richardson-Lucy Deblurring for Scenes under a Projective Motion Path (2011, unknown): “When addressing moving objects, the input image can be segmented into multiple regions each with a constant PSF as demonstrated by Levin [15], Bardsley et al. [2], Cho et al. [7] and Li et al. [18].”
Analyzing spatially-varying blur (2010, background): “More related to our work is the pioneering effort of Levin [15], who also considers single- image input without pre-matting.”
Maintaining Natural Image Statistics with the Contextual Loss (2018, background): “s, balanced training is dicult to achieve and depends on the architecture [8,12]. Well before the CNN era, image priors on the likelihood of patches have been successfully used for image restoration [2,5,13]. Zoran and Weiss [5] showed that constructing a prior from an external corpus of images, and”
Blind motion deblurring from a single image using sparse approximation (2009, background): “[15], Levin [18], Joshi [17]) or manually selecting blurred edges to obtain the local blur kernel (Jia [16]).”
Fast Image Deconvolution using Hyper-Laplacian Priors (2009, background): “1 Related Work Hyper-Laplacian image priors have been used in a range of set tings: super-resolution [20], transparency separation [11] and motion deblurring [9].”
Understanding and evaluating blind deconvolution algorithms (2009, methodology): “However, a direct application of this principle has not yielded the expected results and all algorithms have required additional components, such as marginalization across all possible images [16, 4, 14], spatially-varying terms [7, 19], or solvers that vary their optimization energ y over time [19]”
Non-uniform Deblurring for Shaken Images (2010, background): “Previous work on non-uniform deblurring has focused on piecewise-uniform blurs arising from multiple moving objects in the scene [7] [12], spatially varying combinations of localized uniform blurs [17] [27], or blur arising from rotations of planar objects in the scene [24].”