Time-invariant radon transform by generalized Fourier slice theorem
Open Access
- 1 January 2017
- journal article
- Published by American Institute of Mathematical Sciences (AIMS) in Inverse Problems and Imaging
- Vol. 11 (3) , 501-519
Abstract
Time-invariant Radon transforms play an important role in many fields of imaging sciences, whereby a function is transformed linearly by integrating it along specific paths, e.g. straight lines, parabolas, etc. In the case of linear Radon transform, the Fourier slice theorem establishes a simple analytic relationship between the 2-D Fourier representation of the function and the 1-D Fourier representation of its Radon transform. However, the theorem can not be utilized for computing the Radon integral along paths other than straight lines. We generalize the Fourier slice theorem to make it applicable to general time-invariant Radon transforms. Specifically, we derive an analytic expression that connects the 1-D Fourier coefficients of the function to the 2-D Fourier coefficients of its general Radon transform. For discrete data, the model coefficients are defined over the data coefficients on non-Cartesian points. It is shown numerically that a simple linear interpolation provide satisfactory results and in this case implementations of both the inverse operator and its adjoint are fast in the sense that they run in $O(N \;\text{log}\; N)$ flops, where $N$ is the maximum number of samples in the data space or model space. These two canonical operators are utilized for efficient implementation of the sparse Radon transform via the split Bregman iterative method. We provide numerical examples showing high-performance of this method for noise attenuation and wavefield separation in seismic data.
Keywords
This publication has 29 references indexed in Scilit:
- Deconvolutive Radon transformGeophysics, 2017
- Non‐convex compressed sensing with frequency mask for seismic data reconstruction and denoisingGeophysical Prospecting, 2014
- A Fast Iterative Shrinkage-Thresholding Algorithm for Linear Inverse ProblemsSIAM Journal on Imaging Sciences, 2009
- A Framework for Discrete Integral Transformations II—The 2D Discrete Radon TransformSIAM Journal on Scientific Computing, 2008
- A Framework for Discrete Integral Transformations I—The Pseudopolar Fourier TransformSIAM Journal on Scientific Computing, 2008
- O(N/sup 3/ log N) backprojection algorithm for the 3-D Radon transformIEEE Transactions on Medical Imaging, 2002
- Atomic Decomposition by Basis PursuitSIAM Journal on Scientific Computing, 1998
- The simplest discrete Radon transform1998
- Conjugate Gradient Methods for Toeplitz SystemsSIAM Review, 1996
- Discrete radon transformIEEE Transactions on Acoustics, Speech, and Signal Processing, 1987