fullseye

vol_bounding_box — 3D domain op

使い方

Tight axis-aligned bounding box of a mask’s foreground, in voxel indices.

Returns (z0, y0, x0, z1, y1, x1) with exclusive upper bounds, i.e. vol[z0:z1, y0:y1, x0:x1] is the smallest sub-volume containing every foreground voxel. margin (an int >= 0) grows the box by that many voxels per side, clipped to the volume — headroom for operators with a spatial footprint (a Gaussian of sigma s needs ~3*s voxels of context). A non-{0, 1} input is thresholded at > 0.5.

An empty mask raises ValueError (fail-closed): there is no box, and silently returning the full volume would defeat the point of cropping.

参考(サンプルデータ・文献)

実行できる例(この op を実際に呼ぶ検証済みサンプル)

型が繋がる次の op(primitive を入力に取れる)

fuse_to_voxel · angle_between_lines · angle_between_planes · angle_line_plane · distance_point_plane · distance_point_line · distance_line_line · distance_segment_segment

同カテゴリ(domain)

vol_reduce_domain · vol_crop_domain · vol_uncrop · vol_tiled_map


Provenance: volops.py — 3D 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)。

© 2026 Kazufumi Furuse — Fullseye operator documentation. Licensed under Apache-2.0.