depth op• Data kinds: lightfield → image2d
• Call: import lightfield; lightfield.lf_epi_slope(lf, *, window=9, min_energy=1e-10) (or opslightfield.get("lf_epi_slope"))
Per-pixel slope from the EPI line orientation (structure tensor, one pass).
A scene point traces a straight line in the epipolar-plane image
(:func:lf_epi), so along that line the intensity is constant:
`E_u + s * E_x = 0`. Accumulating that constraint over the whole angular
grid and a `window x window` spatial neighbourhood gives the closed-form
least-squares slope `s = -(J_ux + J_vy) / (J_xx + J_yy)` with
`J_ab = sum(E_a * E_b)` — one pass over the light field, no sweep, both
the horizontal and vertical EPI directions pooled.
This estimator is biased, and the bias is the reason to also run
:func:lf_depth_from_focus. It is ordinary (not total) least squares on
finite differences, so it needs the EPI line to advance less than roughly
one texture correlation length per view. Measured 2026-09-01 on
5x5x64x64 synthetic fields, median over the interior: with texture
`sigma = 1.5 px, true +1.00 -> +1.0004, +0.50 -> +0.5285`,
`+1.50 -> +1.3018, +2.00 -> +1.4614; with sigma = 5.0` px the same
slopes give `+1.0003, +0.5029, +1.4827, +1.9482`. Integer
slopes on a wrapped field come back within 4e-4 and `s = 0` is exact;
`|s| > 1 is under-estimated, by 27% at s = 2` on the roughest texture.
Use it as a fast dense initialiser, not as the final word.
Returns `(slope_map, energy): the (H, W) slope map and the (H, W)`
gradient energy `J_xx + J_yy` that was the denominator. Pixels whose
energy is below *min_energy* have no measurable parallax (a flat patch
of sky); their slope is set to 0 and their energy reported as-is, so you
threshold on `energy` instead of being handed a plausible-looking number
divided by ~0.
Raises `ValueError`: *lf* not a valid light field, an angular/spatial
shape where *neither* EPI direction carries information (the horizontal EPI
needs `U >= 2 **and** W >= 2, the vertical needs V >= 2` and
`H >= 2`), an even or non-positive *window*, a non-positive *min_energy*.
• lightfield_depth family guide
• 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.
• lightfield_depth — py -3.11 examples/lightfield_depth.py
• poc_lightfield_depth — py -3.11 examples/poc_lightfield_depth.py
image2d as input)lf_from_mla · lf_disparity_to_depth · lf_all_in_focus
depth)lf_depth_from_focus · lf_disparity_to_depth · lf_all_in_focus · lf_plenoptic_design
*Provenance: lightfield.py — LIGHTFIELD 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.