features_match — MATCHREG `featmatch` op

• Data kinds: table × table → table

• Call: import fullseye as fs; fs.ledger.features_match(features1, features2, ratio=0.8, mutual=True, max_distance=None) (to call the implementation directly, import featmatch; featmatch.features_match(features1, features2, ratio=0.8, mutual=True, max_distance=None); from the registry, opsmatchreg.get("features_match"))

• GPU: none (CPU reference implementation only)

Usage

Matches two feature tables by nearest neighbour + ratio test (+ mutual nearest neighbour). → table (kind="feature_matches").

> The detailed description below is the original text — the summary and the headings are translated.

記述子が実数(SIFT)なら L2、2 値(ORB)なら Hamming。1 番目に近い相手が 2 番目より明らかに近い(L2 は d1 < ratio · d2、2 乗距離では ratio² で比べる。Hamming は d1 < ratio · d2) ときだけ残し、mutual=True なら相手の側から見ても最近傍であることを求める。

実数の記述子での規則は features.match_descriptors と同じ(同じ入力で同じ組が返ることを tests/test_featmatch.py が固定している)。違いは距離行列を一括で作ること(GPU の影の経路に乗る形)と、2 値の記述子を受けること。

Args:

features1, features2: features_sift / features_orb の表。記述子の種類が同じであること。

ratio: 比の検定のしきい値(0 < ratio <= 1、既定 0.8)。

mutual: 相互最近傍を求めるか。

max_distance: 距離の上限(L2 は距離そのもの、Hamming はビット数)。None なら無し。

Returns:

dict(kind="feature_matches")。idx (M,2) / pts1 / pts2 (M,2) の (x, y) /

distance (M,)(L2 は距離、Hamming はビット数)/ ratio_value (M,)(d1/d2)/ metric /

n1 / n2 / image_shape1 / image_shape2。

Raises:

ValueError: 表の種類が違う / 記述子の種類か次元が違う / 引数が値域外。

Detailed usage guide

• matching_registration family guide

References (sample data, literature)

• 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.

Runnable examples (verified samples that actually call this op)

• poc_part_locate_features — py -3.11 examples/poc_part_locate_features.py

Ops the type connects to (they accept `table` as input)

ransac_homography · ransac_fundamental · registration_to_flow2d

Same category (`featmatch`)

features_sift · features_orb · ransac_homography · ransac_fundamental · features_accuracy_table

*Provenance: featmatch.py — MATCHREG operator registry. This per-op note is generated by tools/opdocs.py md (do not hand-edit).*

features_match — MATCHREG featmatch op