graph op• Data kinds: image2d → table
• Call: import fullseye as fs; fs.ledger.snic_superpixels(image, *, n_segments: 'int' = 100, compactness: 'float' = 0.1, connectivity: 'int' = 4) -> 'Dict[str, object]' (to call the implementation directly, import seggraph; seggraph.snic_superpixels(image, *, n_segments: 'int' = 100, compactness: 'float' = 0.1, connectivity: 'int' = 4) -> 'Dict[str, object]'; from the registry, opssegmentation.get("snic_superpixels"))
SNIC (Algorithm 1 of Achanta–Süsstrunk 2017): non-iterative superpixels grown from grid seeds with a priority queue (snic_superpixels).
> The detailed description below is the original text — the summary and the headings are translated.
距離 d² = |Δx|²/s² + |Δc|²/m²(s = √(N/K)、m = `compactness`、c は画素の値。SLIC の正規化の形、式 (1) の表記は要確認)。
待ち行列から最小の d を取り出し、未ラベルならラベルを付け、重心(位置と値)をその場で更新し、未ラベルの近傍を今の重心との
距離で積む。同点は (距離, 積んだ順) で決める(決定的)。
返す: `labels(1..K)、n_segments、all_connected`(全ラベルが 4-連結 = 構成上の保証を数えて確かめた印)、
`n_disconnected、sizes、seeds((K, 2) の [row, col])、s`。
• halcon_segmentation family guide
• graph_hierarchy_and_thresholds — グラフ・階層・閾値の定理でつくる分割 — 何が厳密で、どこで割れ方が倒れるか
• 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.
• The canonical algorithm (author, year) and its uses are named in the family usage guide above.
• poc_graph_hierarchy_segmentation — py -3.11 examples/poc_graph_hierarchy_segmentation.py
table as input)graph)graph_cut_binary · alpha_expansion · statistical_region_merging · max_tree · area_opening_attr · quasi_flat_zones · alpha_tree · hierarchical_watershed
*Provenance: seggraph.py — SEGMENTATION 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.