by Jean-luc Starck, Emmanuel J. C, David L. Donoho, A. Wavelet, Image Denoising
IEEE Transactions on Image Processing
http://jstarck.free.fr/IEEE02.pdf
Add To MetaCart
Abstract:
Abstract—We describe approximate digital implementations of two new mathematical transforms, namely, the ridgelet transform [2] and the curvelet transform [6], [5]. Our implementations offer exact reconstruction, stability against perturbations, ease of implementation, and low computational complexity. A central tool is Fourier-domain computation of an approximate digital Radon transform. We introduce a very simple interpolation in Fourier space which takes Cartesian samples and yields samples on a rectopolar grid, which is a pseudo-polar sampling set based on a concentric squares geometry. Despite the crudeness of our interpolation, the visual performance is surprisingly good. Our ridgelet transform applies to the Radon transform a special overcomplete wavelet pyramid whose wavelets have compact support in the frequency domain. Our curvelet transform uses our ridgelet transform as a component step, and implements curvelet subbands using a filter bank of à trous wavelet filters. Our philosophy throughout is that transforms should be overcomplete, rather than critically sampled. We apply these digital transforms to the denoising of some standard images embedded in white noise. In the tests reported here, simple thresholding of the curvelet coefficients is very competitive with “state of the art ” techniques based on wavelets, including thresholding of decimated or undecimated wavelet transforms and also including tree-based Bayesian posterior mean methods. Moreover, the curvelet reconstructions exhibit higher perceptual quality than wavelet-based reconstructions, offering visually sharper images and, in particular, higher quality recovery of edges and of faint linear and curvilinear features. Existing theory for curvelet and ridgelet transforms suggests that these new approaches can outperform wavelet methods in certain image reconstruction problems. The empirical results reported here are in encouraging agreement. Index Terms—Curvelets, discrete wavelet transform, FFT, filtering, FWT, radon transform, ridgelets, thresholding rules, wavelets.
Citations
|
196
|
Wavelet-based statistical signal processing using Hidden Markov Models
– Crouse, Nowak, et al.
- 1998
|
|
95
|
Translation invariant de-noising
– Coifman, Donoho
- 1995
|
|
91
|
The Radon Transform and Some of Its Application
– Deans
- 1993
|
|
59
|
Bayesian denoising of visual images in the wavelet domain
– Simoncelli
- 1999
|
|
57
|
Image Processing and Data Analysis: the Multiscale Approach
– STARCK, MURTAGH, et al.
- 1998
|
|
57
|
Wedgelets: Nearly-minimax estimation of edges
– Donoho
- 1999
|
|
54
|
Ridgelets: a key to higher-dimensional intermittency
– Candès, Donoho
- 1999
|
|
37
|
Orthonormal ridgelets and linear singularities
– Donoho
- 2000
|
|
28
|
Wavelet Localization of the Radon Transform
– Olson, DeStefano
- 1994
|
|
26
|
The transformation of poisson, binomial and negative-binomial data
– Anscombe
- 1948
|
|
24
|
Digital Curvelet Transform: Strategy, Implementation, Experiments
– Donoho, Duncan
- 2000
|
|
22
|
Digital reconstruction of multi-dimensional signals from their projections
– Mersereau, Oppenheim
- 1974
|
|
16
|
A New Entropy Measure Based on the Wavelet Transform and Noise Modeling
– Starck, Murtagh, et al.
- 1998
|
|
12
|
Harmonic analysis of neural netwoks
– Candès
- 1999
|
|
11
|
Wavelet sampling and localization schemes for the Radon transform in two dimensions
– Zhao, Welland, et al.
- 1997
|
|
10
|
Orthonormal finite ridgelet transform for image compression
– Do, Vetterli
- 2000
|
|
10
|
Multiscale Entropy Filtering
– Starck, Murtagh
- 1999
|
|
9
|
On the representation of mutilated Sobolev functions
– Candes
- 1999
|
|
8
|
Monoscale ridgelets for the representation of images with edges
– Candes
- 1999
|
|
7
|
Image Reconstruction by the Wavelet Transform Applied to Aperture Synthesis
– Starck, Bijaoui, et al.
- 1994
|
|
6
|
Shiftable Multi-Scale Transforms [or What's Wrong with Orthonormal Wavelets
– Simoncelli, Freeman, et al.
- 1992
|
|
6
|
The wavelet X-ray transform
– Zuidwijk
- 1997
|
|
4
|
Fast ridgelet transforms in dimension 2
– Donoho
- 1997
|
|
3
|
Edge-preserving denoising in linear inverse problems: Optimality of curvelet frames
– Candès, Donoho
- 2000
|
|
3
|
Digital ridgelet transform via rectopolar coordinate transform
– Donoho
- 1998
|
|
3
|
Iterative inversion of the radon transform using imageadaptive wavelet constraints
– Sahiner, Yagle
- 1998
|
|
2
|
The key to higher-dimensional intermittency
– “Ridgelets
- 1999
|
|
2
|
Nearly-minimax estimation of edges
– “Wedgelets
- 1999
|
|
2
|
On the use of wavelets in inverting the Radon transform
– Sahiner, Yagle
- 1992
|
|
2
|
de Zeeuw. The fast wavelet X-ray transform
– Zuidwijk, Paul
- 1999
|
|
1
|
Polar FFT, rectopolar FFT, and applications
– Averbuch, Coifman, et al.
- 2000
|
|
1
|
the representation of mutilated Sobolev functions
– “On
- 1999
|
|
1
|
surprisingly effective nonadaptive representation for objects with edges,” in Curve and Surface Fitting: Saint-Malo
– “Curvelets—A
- 1999
|
|
1
|
ridgelet transform via rectopolar coordinate transform
– “Digital
- 1998
|
|
1
|
ridgelets and linear singularities
– “Orthonormal
- 2000
|
|
1
|
Curvelets–asurprisinglyeffectivenonadaptive representation for objects with edges
– Donoho
- 1999
|
