Current browse context:
math.OC
Change to browse by:
References & Citations
Mathematics > Optimization and Control
Title: Sparse Wavelet Representations of Spatially Varying Blurring Operators
(Submitted on 3 Apr 2014 (v1), last revised 11 Oct 2015 (this version, v5))
Abstract: Restoring images degraded by spatially varying blur is a problem encountered in many disciplines such as astrophysics, computer vision or biomedical imaging. One of the main challenges to perform this task is to design efficient numerical algorithms to approximate integral operators.We introduce a new method based on a sparse approximation of the blurring operator in the wavelet domain. This method requires $\mathcal{O}\left(N \epsilon^{-d/M}\right)$ operations to provide $\epsilon$-approximations, where $N$ is the number of pixels of a $d$-dimensional image and $M\geq 1$ is a scalar describing the regularity of the blur kernel. In addition, we propose original methods to define sparsity patterns when only the operators regularity is known.Numerical experiments reveal that our algorithm provides a significant improvement compared to standard methods based on windowed convolutions.
Submission history
From: Paul Escande [view email][v1] Thu, 3 Apr 2014 17:57:00 GMT (2519kb,D)
[v2] Tue, 15 Apr 2014 17:15:20 GMT (2549kb,D)
[v3] Thu, 22 Jan 2015 19:39:49 GMT (5258kb,D)
[v4] Fri, 23 Jan 2015 14:49:48 GMT (5205kb,D)
[v5] Sun, 11 Oct 2015 10:46:06 GMT (5280kb,D)
Link back to: arXiv, form interface, contact.