dtof op• Datenarten: depth → histcube
• Aufruf: import photoncount; photoncount.dtof_cube_simulate(depth, bins=256, bin_ps=100.0, reflectivity=None, signal_photons=20.0, ambient_photons=5.0, irf_fwhm_ps=200.0, seed=0, noise=True) (oder opsphoton.get("dtof_cube_simulate"))
Synthetisiert den `(H, W, T)`-Photonenhistogrammwürfel eines SPAD-Arrays.
> Die ausführliche Beschreibung unten ist der Originaltext — Zusammenfassung und Überschriften sind übersetzt.
The per-pixel version of :func:tcspc_simulate: every pixel of the *depth*
map (metres, one-way distance) gets a Gaussian return at its own round-trip
time `2d/c`, scaled by *signal_photons* times that pixel's *reflectivity*,
on a uniform ambient pedestal of `ambient_photons/bins` per bin, Poisson
sampled with `numpy.random.default_rng(seed)`.
The output is the cube that :func:dtof_cube_depth inverts, and the axis
order is (H, W, T) with time LAST — the same layout a SPAD array streams.
That is not the `(D, H, W) of a :mod:volops` voxel volume; the two are
both 3-D float arrays and swapping them silently produces a plausible-wrong
depth map, which is why :func:dtof_cube_depth checks and says so.
`noise=False` returns the exact expectation cube (no sampling).
Ground truth: with `noise=False` the per-pixel centroid of the cube returns
the input depth map to an RMS error of 3.2e-16 m (pinned in the tests) — the
pulse integral is analytic, so the only error is float round-off.
Raises `ValueError`: a non-2-D, non-finite or non-positive *depth*, a
*reflectivity* that is negative or not the same shape as *depth*, a *bins*
outside `[2, MAX_BINS]`, non-positive *bin_ps* / *irf_fwhm_ps*, negative
photon budgets, a non-integer *seed*, a cube over
:data:MAX_CUBE_ELEMENTS (`H*W*bins` grows fast — 512x512x256 is 8x the
cap), and any depth whose round-trip time falls outside the time window
(which a real sensor would alias into a short distance).
• Leitfaden zur Familie photon_timeresolved
• Katalog der Beispieldaten (Download-URLs / Lizenzen) — 2-D nutzt skimage.data (BSD/Public Domain) plus synthetische Bilder, 3-D nennt Download-URLs echter Datenquellen (Stanford, PDS, …).
• Herkunft und Literatur der Operatoren — die Quellen der Forschung/Verfahren, auf denen diese Operatorfamilie beruht.
• Der kanonische Algorithmus (Autor, Jahr) und seine Anwendungen stehen im Familienleitfaden oben.
• photon_timeresolved — py -3.11 examples/photon_timeresolved.py
• poc_dtof_ranging — py -3.11 examples/poc_dtof_ranging.py
histcube als Eingabe)dtof)*Provenance: photoncount.py — PHOTON Operator-Registry. Diese Notiz wird von tools/opdocs.py md erzeugt (nicht von Hand bearbeiten).*
© 2026 Kazufumi Furuse — Fullseye operator documentation. Licensed under Apache-2.0.