points_lepetit — MV `detect` op

• Data kinds: image2d → keypoints

• Call: import fullseye as fs; fs.ledger.points_lepetit(image, radius=3, check_neighbor=1, min_check_neighbor_diff=0.06, min_score=0.2, subpix=True) (to call the implementation directly, import mvimage; mvimage.points_lepetit(image, radius=3, check_neighbor=1, min_check_neighbor_diff=0.06, min_score=0.2, subpix=True); from the registry, opsmv.get("points_lepetit"))

• HALCON equivalent: points_lepetit (the HALCON reference is a useful guide to its meaning and parameters)

• GPU: none (CPU reference implementation only)

Usage

Detect interest points by comparing opposite pairs on a circle (the Lepetit-Fua detector).

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

中心の灰色値に近い向かい合う 2 点が円周のどこかにあれば、そこを通る直線が伸びていると見なして落とす。残った画素を、円周の 2 階差分の総和(離散ラプラシアンの近似)で採点し、半径 radius の窓の極大だけを残す。subpix が真なら応答の放物線で小数の位置へ寄せる。灰色の範囲は [0, 1] を想定するので、min_check_neighbor_diff と min_score はその尺度で渡す。平らな画像は空の点列を返す。空の入力(image が 0 要素)は ValueError。

Args:

image: (H, W) の画像。

radius: 比較する円の半径(画素、>= 1)。

check_neighbor: 円周で左右に何点まで一緒に比べるか(>= 0)。

min_check_neighbor_diff: 「近い」と見なす灰色の差の上限。

min_score: 応答(円周平均の 2 階差分の絶対値)の下限。

subpix: 位置を小数へ寄せるか。

Returns:

(N, 2) の (row, col)。応答の強い順。

出典: V. Lepetit and P. Fua, IEEE Trans. PAMI 28(9), 1465-1479 (2006), doi:10.1109/TPAMI.2006.188。

Detailed usage guide

• machine_vision_classics 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)

• machine_vision_tour — py -3.11 examples/machine_vision_tour.py

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

get_grayval_contour_xld · get_circle_pose · gen_nurbs_interp

Same category (`detect`)

create_ncc_model · find_ncc_models

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

points_lepetit — MV detect op