layout op• Data kinds: none → signal (an op determined by its arguments alone — it takes no image or data input)
• Call: import tomography; tomography.projection_angles(n_angles=180, span_deg=180.0, scheme='uniform', start_deg=0.0) (or opstomography.get("projection_angles"))
The angle sequence of a scan, in degrees, as a 1-D float64 array.
Three schemes, and the difference between them is what happens when a scan is
cut short:
• `"uniform" — start + span * k / n`. The textbook scan. A prefix of it
covers only a wedge, so an interrupted uniform scan is a limited-angle scan.
• `"golden" — increments of 180/phi = 111.246...` degrees, wrapped into
`[start, start+span)`. Every prefix is near-uniform, so an interrupted
golden scan is a *sparse* scan, which is a far easier problem. Largest
angular gap left by a scan that stops early, measured:
scheme after 32 of 180 all 180
uniform 149.000 deg 1.000 deg
golden 10.031 deg 1.464 deg
bit-reversed 8.000 deg 1.000 deg
Uniform's 149-degree hole is the entire limited-angle problem arriving by
accident. Bit-reversed is the only one of the three that is good at both
ends; golden's price for working at *every* prefix length rather than only
at powers of two is a completed set 1.46x less even than the grid.
• `"bit-reversed"` — the uniform grid, visited in bit-reversed order. Same
guarantee as golden for power-of-two prefixes and exactly uniform at the
end, which golden is not.
*span_deg* is the total angular range. 180 degrees is the complete data set
for parallel beam — projections at `theta and theta+180` are mirror
images and carry no new information — so a 360-degree span is redundancy, not
resolution, and anything under 180 is the limited-angle problem.
:param n_angles: number of views, `1 .. 65536`.
:param span_deg: total angular range in degrees, `> 0`.
:param scheme: one of :data:ANGLE_SCHEMES.
:param start_deg: angle of the first view.
:returns: `(n_angles,)` float64 array of degrees.
:raises ValueError: on a non-int count, a non-positive span, a non-finite
start, or an unknown scheme.
• 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
• tomography_reconstruct — py -3.11 examples/tomography_reconstruct.py
signal as input)—
layout)*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.