decode op• 資料種類:lightfield → image2d
• 呼叫:import lightfield; lightfield.lf_to_mla(lf)(或 opslightfield.get("lf_to_mla"))
把光場重新交織成微透鏡陣列的原始影格(精確逆變換)。
> 以下的詳細說明為原文 —— 摘要與標題已翻譯。
Inverse of :func:lf_from_mla with `offset=(0, 0)`: the returned frame has
shape `(H*V, W*U) and puts L[v, u, t, s]` back at raw pixel
`(t*V + v, s*U + u). lf_from_mla(lf_to_mla(L), (V, U)) returns L`
bit-for-bit (verified with `np.array_equal`), which is the cheapest
possible check that the decode's index arithmetic has no off-by-one.
Raises `ValueError`: *lf* not 4-D / non-finite / over the shape and
element caps (see :func:lf_from_mla), and a raw frame whose side would
exceed `MAX_SPATIAL * MAX_ANGULAR`.
• 範例資料目錄(下載 URL / 授權) —— 2-D 用 skimage.data(BSD/公有領域)加合成圖,3-D 給出真實資料源(Stanford/PDS 等)的下載 URL。
• 運算子來歷與參考文獻 —— 該運算子族所依據的研究/方法出處。
• 演算法的正典(作者・年份)與用途見上面的族使用指南。
• lightfield_depth — py -3.11 examples/lightfield_depth.py
image2d 作為輸入)lf_from_mla · lf_disparity_to_depth · lf_all_in_focus
decode)*Provenance: lightfield.py — LIGHTFIELD 運算子登記表。本條目由 tools/opdocs.py md 自動產生(請勿手動編輯)。*
© 2026 Kazufumi Furuse — Fullseye operator documentation. Licensed under Apache-2.0.