Scope reduction: remove rail/3D, focus on portal-frame pole 2D detection

- Remove Module B (rail segmentation) and Module C (2D→3D triangulation)
- Rename project: RailPose3D → PoleDetect2D
- Update keypoint schema: {base,top,L_arm,R_arm} → {foot_L,foot_R,head_L,head_R}
- Sprint table reduced from 9 to 5: S0–S4 (pole-only)
- Mark rail-detector-builder and triangulation-builder as INACTIVE
- SfM poses kept for self-training pseudo-label generation only (no 3D output)

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

.claude/agents/pole-detector-builder.md

@@ -1,6 +1,6 @@
 ---
 name: pole-detector-builder
-description: RailPose3D Module A (전철주 4-keypoint pose detection)의 Generator 에이전트. RTMPose-m/ViTPose 학습·추론·copy-paste augmentation·Grounding-DINO+SAM2 zero-shot 부트스트랩을 구현. Module A 관련 sprint(S1 부분, S2, S6의 A 부분, S8) 에서 호출.
+description: PoleDetect2D Module A (라멘구조 전철주 4-keypoint pose detection)의 Generator 에이전트. RTMPose-m/ViTPose 학습·추론·copy-paste augmentation·Grounding-DINO+SAM2 zero-shot 부트스트랩을 구현. Module A 관련 sprint(S1, S2, S3의 A 부분, S4) 에서 호출.
 model: inherit
 tools: Read, Write, Edit, Glob, Grep, Bash
 color: green
@@ -18,8 +18,10 @@ color: green
 ## 기술 스택 (CLAUDE.md 와 일치 필수)
 
 - **모델**: RTMPose-m (primary) → ViTPose-Base (fallback). YOLO11-pose 도 옵션.
-- **Keypoint schema**: COCO-keypoints 포맷, 4점 `{base, top, left_crossarm_tip, right_crossarm_tip}` per pole.
-- **Bootstrap**: Grounding-DINO 1.5 (`"utility pole"`, `"전봇대"`) → SAM 2.1 box prompt → 마스크 PCA → base 후보.
+- **Keypoint schema**: COCO-keypoints 포맷, 4점 `{foot_L, foot_R, head_L, head_R}` per 라멘구조 전철주.
+  - foot_L/R: 왼쪽/오른쪽 기둥 하단 (지면 접점)
+  - head_L/R: 왼쪽/오른쪽 기둥 상단 (가로 보 연결점)
+- **Bootstrap**: Grounding-DINO 1.5 (`"portal frame pole"`, `"전철주"`, `"문형 가구"`) → SAM 2.1 box prompt → 마스크에서 좌우 기둥 PCA → foot/head 후보.
 - **Augmentation**: Albumentations + copy-paste (SAM2 mask 기반, keypoint 보존 변환).
 - **Self-training**: SfM 카메라 포즈 reprojection 으로 pseudo-label 생성, reprojection error > 5px outlier 제거.