forward op• Data kinds: image2d → sinogram
• Call: import tomography; tomography.radon_transform(image, angles_deg=None, n_detectors=None, oversample=1) (or opstomography.get("radon_transform"))
Forward parallel-beam projection: a slice in, a sinogram out.
Rows of the result are projection angles and columns are detector bins. The
convention is fixed here and never negotiated again; every other operator in
this module reads it the same way, and a transposed sinogram is structurally
indistinguishable from a valid one (see :mod:opstomography for the measured
consequences of not giving it its own sort).
The ray at detector bin `j and angle theta` is the line
`x cos(theta) + y sin(theta) = j - (n_det-1)/2, with x` the column
offset from the image centre and `y the row offset — so +y` runs *down*
the array, matching the rest of Fullseye's image indexing rather than a
textbook's upward `y`. The transform is the same either way (the sinogram is
mirrored in the angle axis), but only one of the two agrees with
:func:ellipse_phantom, and the tests hold them together.
Accuracy against the closed form (a disc of radius 60 px in a 256-px grid,
180 views), measured in `tests/test_tomography.py`: interior RMS error
0.073 % of the peak line integral, whole-sinogram RMS 0.402 % — the
difference between the two being the partial-volume edge, where the phantom's
own anti-aliased boundary is what is being sampled.
:param image: 2-D slice, at least 2x2.
:param angles_deg: 1-D view angles in degrees; `None` ->
`linspace(0, 180, 180, endpoint=False)`.
:param n_detectors: bins; `None` -> odd count covering the diagonal.
:param oversample: ray samples per pixel, `1 .. 8`. The default is 1
because 4 measures no better (0.073 % against 0.070 %).
:returns: `(n_angles, n_detectors)` float64 sinogram.
:raises ValueError: on non-finite input, an empty angle list, a detector count
under 4, or a sinogram over :data:MAX_SINOGRAM_ELEMENTS.
• Sample-data catalog (download URLs / licences) — 2-D uses skimage.data (BSD/public domain) plus synthetic images; 3-D lists download URLs for real data sources (Stanford, PDS, …).
• Operator provenance and references — the sources of the research/methods this op family came from.
• The canonical algorithm (author, year) and its uses are named in the family usage guide above.
• ct_reconstruction — py -3.11 examples/ct_reconstruction.py
• poc_ct_fidelity — py -3.11 examples/poc_ct_fidelity.py
sinogram as input)backproject_sinogram · filtered_backprojection · sart_reconstruct · beam_hardening_apply · beam_hardening_correct · ring_artifact_apply · ring_artifact_remove · metal_trace_interpolate
forward)ellipse_phantom · ellipse_sinogram
*Provenance: tomography.py — TOMOGRAPHY operator registry. This per-op note is generated by tools/opdocs.py md (do not hand-edit).*
© 2026 Kazufumi Furuse — Fullseye operator documentation. Licensed under Apache-2.0.