rle_region op• Data kinds: rle_region × rle_region → rle_region
• Call: import fullseye as fs; fs.ledger.vol_rle_union(a, b) (to call the implementation directly, import volregion; volregion.vol_rle_union(a, b); from the registry, ops3d.get("vol_rle_union"))
Union of two RLE regions, computed on the runs (no decode). Cost scales
> The detailed description below is the original text — the summary and the headings are translated.
with the run counts, not the voxel counts — merging two 512**3 masks never
touches 512**3 anything. Regions must share the same volume shape.
手順(`_rle_boolean 共通): 両 region の run を、平面行ごとに W + 1` の
stride を取った 1 本の整数直線に写し(+1 の隙間で隣の行の run が結合しない)、
各 run の start/end をイベントとして並べ、区間ごとの被覆状態 `(a の内側, b の内側)`
を掃引して `ia | ib` が真の区間を最大長の run にまとめ直す。計算量は run 数
`n に対して O(n log n)(np.unique` のソート)。
返り値: 同じ `shape の新しい VolRLE`(run は行順・x 昇順、隣接・重複する run は
1 本に併合済み)。両方が空なら run 0 本の region。
`vol_rle_decode(union) == decode(a) | decode(b)` が voxel 単位で成り立つ。
検証(`ValueError): どちらかが VolRLE` でない・整合性検査に失敗 /
`a.shape != b.shape`(別の volume に住む region は合成できない)。
• 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.
• rle_region_efficiency — py -3.11 examples_3d/rle_region_efficiency.py
rle_region as input)vol_rle_decode · vol_rle_volume · vol_rle_bbox · vol_rle_centroid · vol_rle_intersect · vol_rle_difference · fuse_to_voxel
rle_region)vol_rle_encode · vol_rle_decode · vol_rle_volume · vol_rle_bbox · vol_rle_centroid · vol_rle_intersect · vol_rle_difference · vol_rle_components
*Provenance: volregion.py — 3D 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.