decode op• Data kinds: lightfield → table
• Call: import lightfield; lightfield.lf_stats(lf) (or opslightfield.get("lf_stats"))
Describe a light field: shape, angular centre, and the slope range it can carry.
Returns a dict — `angular_v / angular_u · height / width` ·
`n_views = V*U · center_v / center_u ((N-1)/2`, a half-integer
when the axis is even, so `center_is_a_view` says whether a single view
actually sits at the centre) · `min / max / mean` of the samples ·
`max_slope_px, the largest |s|` for which the extreme view is still
displaced by less than half the frame, i.e. the honest limit of what this
array can measure (`min(H, W) / max(V-1, U-1, 1)`; reported as
`float(min(H, W))` when the array has a single view in both directions,
since then no shift happens at all) · `baseline_views`, the angular span
`(V-1, U-1)` that any disparity is measured over.
Raises `ValueError`: *lf* is not a finite 4-D array within the shape
and element caps.
• 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
table as input)—
decode)*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.