lf_from_mla — LIGHTFIELD `decode` op

• データ種: image2d → lightfield

• 呼び出し: import lightfield; lightfield.lf_from_mla(raw, angular=(5, 5), *, offset=(0, 0), crop=False) (または opslightfield.get("lf_from_mla"))

Decode a microlens-array raw frame into a `(V, U, H, W)` light field.

The plenoptic sensor stores, for microlens `(t, s), a V x U` block of

pixels — one per direction through the main lens' exit pupil. So raw pixel

`(offset_y + t*V + v, offset_x + s*U + u) is exactly L[v, u, t, s]`:

the decode is a pure re-sort, no interpolation, no data invented.

:func:lf_to_mla inverts it bit-exactly.

*offset* is the integer position of the first whole microlens; a real MLA is

never aligned to pixel `(0, 0)`. Sub-pixel MLA calibration is out of scope

(see the module docstring) — this operator works on an already-rectified

frame.

By default a raw frame whose usable size is not a whole multiple of the

microlens pitch is a `ValueError`, because the alternative — quietly

dropping the last partial row of microlenses — shifts every subsequent

microlens centre and produces a light field that looks right and is wrong.

Pass `crop=True` to opt in to that crop explicitly.

Raises `ValueError`: *raw* not 2-D or non-finite, *angular* outside

`[1, MAX_ANGULAR]`, a negative *offset*, an *offset* that leaves fewer

than one whole microlens, a size that is not a multiple of the pitch (unless

`crop=True), a decoded sub-aperture image over :data:MAX_SPATIAL` on a

side, and a decoded size over :data:MAX_LF_ELEMENTS.

• サンプルデータカタログ(DL URL / ライセンス) — 2-D は skimage.data(BSD/public)+ 合成、3-D は実データ源(Stanford/PDS 等)の DL URL。

• 演算子の来歴・参考文献 — この op 族の元になった研究/手法の出典。

• アルゴリズムの正典(著者・年)と用途は上記ファミリ使い方ガイドに記載。

• lightfield_depth — py -3.11 examples/lightfield_depth.py

*Provenance: lightfield.py — LIGHTFIELD operator registry. この per-op ノートは tools/opdocs.py md が自動生成(手編集しない)。*

lf_from_mla — LIGHTFIELD decode op