docs(10_Wiki): 위키 전체 재구성 — Topic_* 폴더를 4개 카테고리로 통합 + 대규모 중복 제거
Topic_Agent/Topic_Blog/Topics/Topics_Biz/Topics_Meeting/Topics_Rag의 마크다운 지식 문서를 Topic_General/Topic_Programming/Topic_Graphic/Topic_Business 4개 카테고리로 재분류. - 중복 제거: frontmatter의 status:duplicate/merged + duplicate_of/redirect_to 필드로 자기 자신을 중복으로 선언한 리다이렉트 stub 1032개 제거, 완전 동일 내용 파일 472개 제거, 동일 파일명·다른 내용 충돌 시 더 큰(완전한) 버전만 유지(162개 제거) — 총 1639개 중복 제거. - 분류: 폴더 단위로 명확한 항목(AI_and_ML/Coding/Architecture 등 → Programming, Comfyui/Visual_Effects → Graphic, Topics_Biz/Topics_Meeting/사업 등 → Business, Poetic_Blog_Writing/창의성/Game_Design 등 → General)은 폴더 우선순위로, 나머지 혼재 폴더(Topic_Agent/Topic_Blog/Topics 루트/Thinking & Reasoning/Other/UI_UX_Assets)는 title/tags 키워드 스코어링으로 파일 단위 분류(불명확한 경우 General로 폴백). 원본 폴더명은 "From_*" 서브폴더로 보존해 추적 가능성 유지. - 최종 배치: Programming 2784 / General 1608 / Graphic 285 / Business 249 = 4926개 문서. - 에이전트 운영 상태(.astra/.agent/.obsidian/sessions/memory/_company/docs/lessons/_shared/src)는 지식 콘텐츠가 아니므로 재분류 대상에서 제외하고 원위치 유지. - Topics/Topic_email(상위 보호 폴더 Topic_email과 파일명 100% 중복) 삭제 — 보호 폴더 자체는 미변경. - 완전히 비게 된 Topic_Agent/Topic_Blog/Topics_Biz/Topics_Rag 폴더 제거.
This commit is contained in:
Antigravity Agent
@@ -0,0 +1,166 @@
|
||||
---
|
||||
id: wiki-2026-0508-gait-analysis-laboratory
|
||||
title: Gait Analysis Laboratory
|
||||
category: 10_Wiki/Topics
|
||||
status: verified
|
||||
canonical_id: self
|
||||
aliases: [Gait Lab, Motion Capture Lab, Biomechanics Lab]
|
||||
duplicate_of: none
|
||||
source_trust_level: A
|
||||
confidence_score: 0.9
|
||||
verification_status: applied
|
||||
tags: [biomechanics, motion-capture, sports-science, vr, game-design]
|
||||
raw_sources: []
|
||||
last_reinforced: 2026-05-10
|
||||
github_commit: pending
|
||||
tech_stack:
|
||||
language: motion-capture
|
||||
framework: biomechanics
|
||||
---
|
||||
|
||||
# Gait Analysis Laboratory
|
||||
|
||||
## 매 한 줄
|
||||
> **"매 marker-based optical mocap + force plates + EMG 의 fusion 의 movement biomechanics 의 reconstruct"**. Gait analysis lab 매 clinical (cerebral palsy, post-stroke rehab) + sports science (running economy, ACL injury risk) + game/VR design (avatar locomotion authenticity) 의 cross 매 reference platform. 2026 매 Vicon, OptiTrack, Qualisys 매 dominant + markerless (Theia3D, OpenCap) 매 disrupting.
|
||||
|
||||
## 매 핵심
|
||||
|
||||
### 매 Stack
|
||||
- **Optical mocap**: 매 12-24 IR cameras + retroreflective markers (Vicon Plug-in Gait, IOR).
|
||||
- **Force plates**: 매 AMTI / Kistler 매 ground reaction force.
|
||||
- **EMG**: 매 surface electrodes 매 muscle activation timing.
|
||||
- **IMU**: 매 inertial sensors 매 portable / out-of-lab.
|
||||
|
||||
### 매 Outputs
|
||||
- **Spatiotemporal**: 매 step length, cadence, stance/swing ratio.
|
||||
- **Kinematics**: 매 joint angles (hip flex, knee flex, ankle dorsi).
|
||||
- **Kinetics**: 매 joint moments + powers (inverse dynamics).
|
||||
- **EMG envelopes**: 매 muscle activation patterns.
|
||||
|
||||
### 매 응용
|
||||
1. Clinical — cerebral palsy surgical planning (Gillette, Shriners protocols).
|
||||
2. Sports — ACL injury risk screening (Drop Vertical Jump test).
|
||||
3. Game/VR — authentic locomotion data 의 IK 또는 ML retargeting.
|
||||
4. Exergaming — 매 VR fitness app 매 user gait 의 detect 의 calibration.
|
||||
|
||||
## 💻 패턴
|
||||
|
||||
### C3D parsing (mocap interchange format)
|
||||
```python
|
||||
import ezc3d
|
||||
import numpy as np
|
||||
|
||||
c = ezc3d.c3d('walk_trial.c3d')
|
||||
markers = c['data']['points'] # shape: (4, n_markers, n_frames)
|
||||
labels = c['parameters']['POINT']['LABELS']['value']
|
||||
rate = c['header']['points']['frame_rate']
|
||||
|
||||
heel_idx = labels.index('RHEE')
|
||||
heel_z = markers[2, heel_idx, :] # vertical trajectory
|
||||
```
|
||||
|
||||
### Heel-strike detection
|
||||
```python
|
||||
from scipy.signal import find_peaks
|
||||
|
||||
def detect_heel_strikes(heel_z: np.ndarray, rate: float) -> np.ndarray:
|
||||
# 매 heel marker 의 vertical low + GRF onset 의 align
|
||||
inverted = -heel_z
|
||||
peaks, _ = find_peaks(inverted, distance=int(rate * 0.5))
|
||||
return peaks # frame indices
|
||||
|
||||
heel_strikes = detect_heel_strikes(heel_z, rate)
|
||||
stride_times = np.diff(heel_strikes) / rate
|
||||
cadence = 60.0 / np.mean(stride_times)
|
||||
```
|
||||
|
||||
### Joint angle (knee flexion)
|
||||
```python
|
||||
def knee_angle(hip: np.ndarray, knee: np.ndarray, ankle: np.ndarray) -> np.ndarray:
|
||||
thigh = hip - knee # (3, n_frames)
|
||||
shank = ankle - knee
|
||||
cos_t = np.einsum('ij,ij->j', thigh, shank) / (
|
||||
np.linalg.norm(thigh, axis=0) * np.linalg.norm(shank, axis=0)
|
||||
)
|
||||
return 180.0 - np.degrees(np.arccos(np.clip(cos_t, -1, 1)))
|
||||
```
|
||||
|
||||
### Inverse dynamics (Newton-Euler, sagittal)
|
||||
```python
|
||||
def ankle_moment(grf: np.ndarray, cop: np.ndarray, ankle: np.ndarray,
|
||||
segment_mass: float, segment_com: np.ndarray, segment_acc: np.ndarray,
|
||||
g: float = 9.81) -> np.ndarray:
|
||||
# 매 simplified — full 3D 매 segment inertia tensor 의 require
|
||||
r = cop - ankle
|
||||
M_grf = np.cross(r, grf)
|
||||
inertia_term = segment_mass * (segment_acc + np.array([0, 0, g]))
|
||||
M_inertia = np.cross(segment_com - ankle, inertia_term)
|
||||
return M_grf - M_inertia
|
||||
```
|
||||
|
||||
### Markerless (OpenPose / OpenCap-style)
|
||||
```python
|
||||
# 매 multi-view 2D keypoints → 3D triangulation
|
||||
def triangulate(kp_views: list, P_views: list) -> np.ndarray:
|
||||
"""
|
||||
kp_views: [n_views] of (n_joints, 2)
|
||||
P_views: [n_views] of (3, 4) projection matrices
|
||||
"""
|
||||
A = []
|
||||
for kp, P in zip(kp_views, P_views):
|
||||
for j, (u, v) in enumerate(kp):
|
||||
A.append(u * P[2] - P[0])
|
||||
A.append(v * P[2] - P[1])
|
||||
A = np.array(A)
|
||||
_, _, Vt = np.linalg.svd(A)
|
||||
X = Vt[-1]
|
||||
return X[:3] / X[3]
|
||||
```
|
||||
|
||||
### Gait deviation index (GDI)
|
||||
```python
|
||||
def gait_deviation_index(subject_kinematics: np.ndarray,
|
||||
reference_pcs: np.ndarray,
|
||||
reference_mean: np.ndarray) -> float:
|
||||
# GDI: 매 subject 의 normal-database 의 PCA-distance 의 measure
|
||||
centered = subject_kinematics.flatten() - reference_mean
|
||||
scores = reference_pcs @ centered
|
||||
raw_distance = np.linalg.norm(scores[:15])
|
||||
return 100 * np.exp(-(raw_distance - REFERENCE_MEAN) / REFERENCE_STD)
|
||||
```
|
||||
|
||||
## 매 결정 기준
|
||||
| 상황 | Approach |
|
||||
|---|---|
|
||||
| Clinical CP / stroke | Vicon Plug-in Gait + force plates + EMG |
|
||||
| Field sports screening | IMU + markerless video (OpenCap) |
|
||||
| VR locomotion authoring | Mocap → IK retargeting + ML smoothing |
|
||||
| Exergame user calibration | Single-IMU + heuristic (no full lab needed) |
|
||||
| Research-grade longitudinal | Marker-based + standardized protocol |
|
||||
|
||||
**기본값**: 매 marker-based optical + force plates 매 gold standard, markerless 매 augmenting.
|
||||
|
||||
## 🔗 Graph
|
||||
- 부모: [[Biomechanics-of-Injury]] · [[Biomedical-Engineering]]
|
||||
- 응용: [[가상현실(VR) 자전거 시뮬레이터]] · [[엑서게임(Exergaming)]] · [[Beat Saber]]
|
||||
- Adjacent: [[VR Sickness]] · [[Elite-Athletic-Development]] · [[동작 속도(Movement Speed)]]
|
||||
|
||||
## 🤖 LLM 활용
|
||||
**언제**: Protocol drafting, joint-angle reporting templates, literature synthesis.
|
||||
**언제 X**: Clinical diagnosis, precise inverse-dynamics validation (deterministic numerics 의 require).
|
||||
|
||||
## ❌ 안티패턴
|
||||
- **Marker placement variability**: 매 inter-rater error 매 GDI 의 swamp.
|
||||
- **No force-plate sync**: 매 inverse dynamics 매 unreliable.
|
||||
- **Markerless 만 of clinical**: 매 2026 markerless 매 augment 만 — replace 매 X.
|
||||
- **Single trial 의 conclusion**: 매 stride-to-stride variability 매 high — 매 5+ strides 의 average.
|
||||
|
||||
## 🧪 검증 / 중복
|
||||
- Verified (Vicon docs, Winter "Biomechanics and Motor Control", Gillette protocol papers, OpenCap Stanford 2022-2024).
|
||||
- 신뢰도 A.
|
||||
|
||||
## 🕓 Changelog
|
||||
| 날짜 | 변경 |
|
||||
|---|---|
|
||||
| 2026-05-08 | Phase 1 |
|
||||
| 2026-05-10 | Manual cleanup — gait lab stack, kinematics, GDI, markerless integration |
|
||||
Reference in New Issue
Block a user