domain op• 資料種類:voxel × voxel → voxel
• 呼叫: import fullseye as fs; fs.ledger.vol_reduce_domain(vol, domain)(要直接呼叫實作,import volops; volops.vol_reduce_domain(vol, domain);從台帳取用則 ops3d.get("vol_reduce_domain"))
把體限制在 *domain* 遮罩內(HALCON `reduce_domain`,逐體素)。
> 以下的詳細說明為原文 —— 摘要與標題已翻譯。
Every voxel outside the foreground of *domain* is set to `0` — the same
"outside the domain is undefined -> 0" convention the 2-D `it_crop_domain`
op uses (a plain numpy array cannot carry a separate domain channel, so the
restriction is materialised). A non-`{0, 1}` *domain* is thresholded at
`> 0.5`. Use it to silence everything a downstream operator must not see
(metal artefacts, the scanner bed, a neighbouring part); combine with
:func:vol_crop_domain when you also want the *memory* of the volume
reduced, not just its values.
Returns a `(D, H, W)` float64 volume of the same shape.
• 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。
• roi_domain_boundary — py -3.11 examples_3d/roi_domain_boundary.py
voxel 作為輸入)voxel_to_mips · voxel_to_mesh · signed_distance_field · to_points · sobel3d · hessian3d · curvature_maps · edt_jfa
domain)vol_bounding_box · vol_crop_domain · vol_uncrop · vol_tiled_map
*Provenance: volops.py — 3D 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.