photon_sample — PHOTON counting op

Data kinds: image2dimage2d

Call: import photoncount; photoncount.photon_sample(image, photons_per_unit=100.0, dark_rate=0.0, seed=0) (or opsphoton.get("photon_sample"))

Usage

Poisson-sample an expected-photon image into an actual photon count image.

*image* is a non-negative 2-D map of scene radiance in arbitrary units;

`lambda = image * photons_per_unit + dark_rate` is the expected number of

photons in each pixel over the exposure, and the result is one Poisson

realisation of it. *dark_rate* is the dark-count contribution (a SPAD counts

thermally generated carriers even in the dark) in the same photon units.

Returns the counts themselves as a float64 `(H, W)` image (integer

valued). That is the deliberate difference from

:func:backends_aug.aug_shot_noise, which returns `Poisson(v*K)/K` clipped

to `[0, 1]` for training-data augmentation: every operator downstream here

(Fano factor, Anscombe, Coates, dToF) needs `N`, and the rescale-and-clip

is not invertible.

`seed` is a required non-negative integer and the RNG is

`numpy.random.default_rng(seed)` — same seed, same frame, on any machine.

Ground truth it reproduces (pinned in `tests/test_photoncount.py`): the

sample mean and sample variance both converge to `lambda`. Measured on a

flat `lambda = 100` field of 512x512 pixels at seed 0 — mean 99.9796,

Fano factor 1.001089, so the photon-limited SNR is `sqrt(lambda)`: 9.9990

predicted from the mean, 9.9935 actually achieved.

Raises `ValueError`: negative or non-finite *image*, negative

*photons_per_unit* / *dark_rate*, a non-integer or negative *seed*, an image

over :data:MAX_IMAGE_ELEMENTS, and — instead of letting numpy fail deep

inside the sampler — any `lambda over :data:MAX_LAMBDA`.

Detailed usage guide

photon_timeresolved family guide

References (sample data, literature)

• 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.

Runnable examples (verified samples that actually call this op)

photon_timeresolvedpy -3.11 examples/photon_timeresolved.py

Ops the type connects to (they accept image2d as input)

photon_statistics · photon_uncertainty · anscombe_transform · anscombe_inverse

Same category (counting)

photon_statistics · photon_uncertainty


*Provenance: photoncount.py — PHOTON 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.