Space-Variant Restoration of Images Degraded by Camera Motion Blur

Šorel, Michal; Flusser, Jan

doi:10.1109/tip.2007.912928

Cited by 86 publications

(58 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Coded aperture method has been proposed for deblurring and decoded images are used to infer pixel-wise depth in a Markov random field [18]. Depth effect on translational blur has also been discussed in [26] where the blur is estimated from the user specified region with approximately constant depth to infer depth. In [31], two blurry images of the same scene are used to infer depth with stereo techniques and further improve the blur removal.…”

Section: Related Workmentioning

confidence: 99%

Joint Depth Estimation and Camera Shake Removal from Single Blurry Image

Liu

Yang

2014

2014 IEEE Conference on Computer Vision and Pattern Recognition

View full text Add to dashboard Cite

Camera shake during exposure time often results in spatially variant blur effect of the image. The non-uniform blur effect is not only caused by the camera motion, but also the depth variation of the scene. The objects close to the camera sensors are likely to appear more blurry than those at a distance in such cases. However, recent non-uniform deblurring methods do not explicitly consider the depth factor or assume fronto-parallel scenes with constant depth for simplicity. While single image non-uniform deblurring is a challenging problem, the blurry results in fact contain depth information which can be exploited. We propose to jointly estimate scene depth and remove non-uniform blur caused by camera motion via exploiting their underlying geometric relationships, with only single blurry image as input. To this end, we present a unified layer-based model for depth-involved deblurring. We provide a novel layerbased solution using matting to partition the layers and an expectation-maximization scheme to solve this problem. This approach largely reduces the number of unknowns and makes the problem tractable. Experiments on challenging examples demonstrate that both depth and camera shake removal can be well addressed within the unified framework.

show abstract

Section: Related Workmentioning

confidence: 99%

Joint Depth Estimation and Camera Shake Removal from Single Blurry Image

Liu

Yang

2014

2014 IEEE Conference on Computer Vision and Pattern Recognition

View full text Add to dashboard Cite

show abstract

“…The minimum is reached for a negative value of s 1 and the same behavior was observed for any pair of blurs h 1 , h 2 . The data term is thus biased towards kernels with small negative values and the unconstrained optimization problem (19) …”

Section: Overestimated Kernel Sizementioning

confidence: 99%

“…This constraint is automatically enforced by the fidelity term 2 k=1 u * h k − g k in (19). If the mean value of the estimated image u is equal to the mean value of g k , then by solving (19) (h-step) we always preserve i h k (i) = 1.…”

Section: Overestimated Kernel Sizementioning

confidence: 99%

See 1 more Smart Citation

Robust Multichannel Blind Deconvolution via Fast Alternating Minimization

Šroubek

Milanfar

2012

IEEE Trans. on Image Process.

166

128

View full text Add to dashboard Cite

Abstract-Blind deconvolution, which comprises simultaneous blur and image estimation, is a strongly ill-posed problem. It is by now well-known that if multiple images of the same scene are acquired, this multichannel blind deconvolution problem is better posed and allows of blur estimation directly from the degrade images. We improve the multichannel idea by adding robustness to noise and stability in the case of large blurs or if the blur size is vastly overestimated. We formulate blind deconvolution as a 1-regularized optimization problem and seek a solution by alternately optimizing with respect to the image and with respect to blurs. Each optimization step is converted to a constrained problem by variable splitting and then addressed with an augmented Lagrangian method, which permits simple and fast implementation in the Fourier domain. Rapid convergence of the proposed method is illustrated on synthetically blurred data. Applicability is also demonstrated on deconvolution of real photos taken by a digital camera.

show abstract

“…Though there is an extensive body of literature on estimating and removing motion blur from traditional images (e.g. [14], [15], [6], [2]), the elements of a traditional camera impose strict limitations on postprocessing.…”

Section: Introductionmentioning

confidence: 99%

Iris capture from moving subjects using a fluttering shutter

McCloskey

Jelı́nek

2010

2010 Fourth IEEE International Conference on Biometrics: Theory, Applications and Systems (BTAS)

View full text Add to dashboard Cite

Abstract-We address the problem of sharp iris image capture from moving subjects for the purpose of biometric identification. We utilize recent research from the field of computational photography, and capture images using the flutter shutter technique of [12]. Instead of capturing an image with traditional motion blur, we open and close the shutter several times during capture in order to effect invertible motion blur in the captured image. After applying automated blur estimation and de-blurring, we employ de-convolution to estimate a sharp image from the captured image. Through black-box testing with an existing iris matcher, we demonstrate the improved utility of these images for biometrics.

show abstract

Space-Variant Restoration of Images Degraded by Camera Motion Blur

Cited by 86 publications

References 30 publications

Joint Depth Estimation and Camera Shake Removal from Single Blurry Image

Joint Depth Estimation and Camera Shake Removal from Single Blurry Image

Robust Multichannel Blind Deconvolution via Fast Alternating Minimization

Iris capture from moving subjects using a fluttering shutter

Contact Info

Product

Resources

About