feature op• 데이터 종류: voxel → voxel
• 호출: import fullseye as fs; fs.ledger.vol_frangi(vol, scales=(1, 2, 3), alpha=0.5, beta=0.5, c=None, black_ridges=False)(구현을 직접 호출하려면 import volops; volops.vol_frangi(vol, scales=(1, 2, 3), alpha=0.5, beta=0.5, c=None, black_ridges=False), 원장에서 가져오려면 ops3d.get("vol_frangi"))
3-D Frangi 관상도 —— 다중 스케일 관상 구조 강조.
> 아래 상세 설명은 원문입니다 —— 요약과 제목은 번역되어 있습니다.
For every `sigma` in *scales* the gamma-normalised Hessian is formed, its
three eigenvalues `|l1| <= |l2| <= |l3|` are taken, and the Frangi response
V = (1 - exp(-Ra2 / 2 alpha2)) * exp(-Rb2 / 2 beta2)
• (1 - exp(-S2 / 2 c2))
is evaluated, where `Ra = |l2|/|l3| (plate vs. line), Rb = |l1|/sqrt|l2 l3|`
(blob deviation) and `S = sqrt(l1**2 + l2**2 + l3**2)` (structure strength).
The response is set to 0 where the contrast polarity is wrong (bright tube:
`l2 > 0 or l3 > 0; black_ridges=True` flips this). The maximum over
scales is taken and the volume is normalised to `[0, 1]`.
Parameters
----------
scales : sequence of float — Gaussian sigmas, in voxels, to bracket the
vessel *radii* of interest (see the module "scale-dependent" limitation).
alpha, beta : sensitivities of the plate- and blob-suppression terms (Frangi's
defaults 0.5).
c : half the maximum Hessian norm `S at each scale when None` (Frangi's
adaptive suggestion); otherwise a fixed structure-strength scale.
black_ridges : `False` (default) enhances *bright* tubes on a dark
background; `True` enhances *dark* tubes on a bright background.
Returns a `(D, H, W) float64 volume in [0, 1]`. Reference: Frangi et al.,
MICCAI 1998.
• 샘플 데이터 카탈로그(DL URL / 라이선스) —— 2-D 는 skimage.data(BSD/public)+ 합성, 3-D 는 실데이터 소스(Stanford/PDS 등)의 DL URL.
• 연산자의 내력·참고문헌 —— 이 연산자 족의 바탕이 된 연구/기법의 출처.
• vessel_metrology — py -3.11 examples_3d/vessel_metrology.py
• volume_downsampling — py -3.11 examples_3d/volume_downsampling.py
voxel 를 입력으로 받는 것)voxel_to_mips · voxel_to_mesh · signed_distance_field · to_points · sobel3d · hessian3d · curvature_maps · edt_jfa
feature)sobel3d · hessian3d · curvature_maps · edt_jfa · vol_local_std · vol_local_thickness · vol_orientation_coherence · vol_euler_number
*Provenance: volops.py — 3D 연산자 레지스트리. 이 op 노트는 tools/opdocs.py md 가 자동 생성합니다(직접 편집하지 마세요).*
© 2026 Kazufumi Furuse — Fullseye operator documentation. Licensed under Apache-2.0.