Explicit Oblique Inverses of the Approximate Discrete Radon Transform

Abstract

The approximate discrete Radon transform (ADRT) is a fast, hierarchical approximation of the Radon transform whose levels have explicit spectral decompositions, but its Moore-Penrose inverse, needed for data outside the range, has no closed form. We construct explicit left inverses of the ADRT, oblique projections onto its range, that are exact for every choice of their parameters and amplify noise by a constant factor relative to the Moore-Penrose inverse. Realized with polynomial filters on the butterfly structure of the levels, they are fixed linear maps of cost O(N^5/2log^3/2N/σ_min) for an N× N image, where σ_min is the smallest singular value of the ADRT; with the measured growth σ_min∼ N^1/4 this is O(N^2.25log^3/2N). A hybrid that keeps the explicit inverse of the cross-quadrant level is cheaper and closer to the Moore-Penrose inverse. In compiled implementations both outperform a polynomial solver of the normal equations from N=256 on, and their conditioning is traced to the aliasing of integer-sampled lines.

Explicit formula for the discrete Laplace transform of the Möbius function, related special functions, and a criterion for the Riemann hypothesis
Sergey Liflandsky
arXiv14 Sep 2026PreprintCC BY 4.0Full textAlgebra & number theoryarXiv
Sparse Lifts of q-Linearized Permutation Polynomials with Explicit Inverses
Yi Li
arXiv12 Oct 2026PreprintarXiv licenceFull textAlgebra & number theoryarXiv
A sharp discrete Sobolev inequality and left inverses of higher order differences
Raul Hindov, Shahaf Nitzan, Jan-Fredrik Olsen +1
arXiv13 Oct 2026PreprintCC BY 4.0Full textAnalysis & PDEsarXiv
Explicit Exponential Sum Estimates and Approximate Functional Equations for the Zeta Function
Natasha Dhiman, Habiba Kadiri, Emily Quesada-Herrera
arXiv1 Sep 2026PreprintarXiv licenceFull textAlgebra & number theoryarXiv
FASS: Discrete Fourier Transform Adapters for Speech Separation With Incremental Learning
Ziye Yang, Xiang Song, Min Zhao +3
IEEE Transactions on Audio, Speech and Language Processing25 Sep 2026Peer-reviewedPublisher termsFull textStatisticsHAL

Cite

BibTeX RIS

BibTeX

@misc{rim2026explicit,
  title = {{Explicit Oblique Inverses of the Approximate Discrete Radon Transform}},
  author = {Donsub Rim},
  howpublished = {arXiv},
  year = {2026},
  month = oct,
  doi = {10.48550/arxiv.2610.23363},
  eprint = {2610.23363},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2610.23363},
}

RIS

TY  - UNPB
TI  - Explicit Oblique Inverses of the Approximate Discrete Radon Transform
AU  - Donsub Rim
PY  - 2026
DA  - 2026/10/16
PB  - arXiv
DO  - 10.48550/arxiv.2610.23363
UR  - https://arxiv.org/abs/2610.23363
AB  - The approximate discrete Radon transform (ADRT) is a fast, hierarchical approximation of the Radon transform whose levels have explicit spectral decompositions, but its Moore-Penrose inverse, needed for data outside the range, has no closed form. We construct explicit left inverses of the ADRT, oblique projections onto its range, that are exact for every choice of their parameters and amplify noise by a constant factor relative to the Moore-Penrose inverse. Realized with polynomial filters on the butterfly structure of the levels, they are fixed linear maps of cost O(N^5/2log^3/2N/σ_min) for an N× N image, where σ_min is the smallest singular value of the ADRT; with the measured growth σ_min∼ N^1/4 this is O(N^2.25log^3/2N). A hybrid that keeps the explicit inverse of the cross-quadrant level is cheaper and closer to the Moore-Penrose inverse. In compiled implementations both outperform a polynomial solver of the normal equations from N=256 on, and their conditioning is traced to the aliasing of integer-sampled lines.
ER  -

Abstract

Related

Cite