change op• Data kinds: image2d × image2d → table
• Call: import fullseye as fs; fs.ledger.radial_reach_filter(background, image, t_p=0.08, agree_thresh=0.625, max_reach=32, min_dirs=4, min_decided_fraction=0.01) -> 'dict' (to call the implementation directly, import codematch; codematch.radial_reach_filter(background, image, t_p=0.08, agree_thresh=0.625, max_reach=32, min_dirs=4, min_decided_fraction=0.01) -> 'dict'; from the registry, opscodematch.get("radial_reach_filter"))
Radial Reach Filter (RRF): compares a background and the current image by the signs of the reaches in 8 directions.
> The detailed description below is the original text — the summary and the headings are translated.
背景 B の画素 p から方向 d_k へ 1, 2, … と進み、|B(p + r d_k) - B(p)| >= t_p になる最初の r を reach r_k(max_reach まで。画像の外に出たらその方向は無効)。符号 b_k = [B(p + r_k d_k) >= B(p)]、現在の画像でも同じ位置の組で b'_k = [I(p + r_k d_k) >= I(p)] を作り、一致した方向の数を RRC とする。有効な方向のうち一致した割合が agree_thresh 未満の画素を前景にする(8 方向が全部有効なら、既定 0.625 は「一致 5 未満」)。
現在の画像に狭義単調増加の明るさの変換を掛けても、結果は 1 ビットも変わらない。影(局所の掛け算)の内側も同じ理由で前景にならない。
fail-closed: 有効な方向が min_dirs 未満の画素は判定しない(decided が False、rrc が -1、前景にもしない)。判定できた画素が min_decided_fraction 未満なら ValueError(背景が平坦すぎて、変化を言えない)。
返り: foreground(bool)、decided(bool)、rrc(一致した方向の数、判定しない画素は -1)、valid_dirs、agree_fraction(判定しない画素は 1.0)、reach((8, H, W)、無効は 0)、decided_fraction、foreground_fraction、refusal_margin(判定できた割合 − min_decided_fraction)。
• robust_code_matching family guide
• 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_code_matching_breakdown — py -3.11 examples/poc_code_matching_breakdown.py
table as input)—
change)—
*Provenance: codematch.py — CODEMATCH 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.