7 Commits

Author SHA1 Message Date
Rain-Bus 90fe705e8f refactor: 迁移布局检测模型从 PicoDet 到 DocLayout-YOLO
- 核心变更:
  - app/services/layout_detector.py: 重写布局检测器,从 PicoDet-S_layout_3cls 迁移到 DocLayout-YOLO (DocStructBench, imgsz=1024)
  - 支持多设备推理 (CPU/CUDA/DirectML/OpenVINO 等),自动探测最优设备
  - 预处理改为 letterbox (保比例缩放+灰边 padding),坐标还原使用 (model_coord - padding) / ratio 公式
  - 后处理解析 YOLOv10 end-to-end 输出 [N,6]=[x1,y1,x2,y2,conf,cls]
  - 类别映射改为按 class name 动态匹配 (figure/figure_group→picture, table/table_group→table)

- 新增文件:
  - scripts/export_doclayout_yolo_onnx.py: DocLayout-YOLO ONNX 导出脚本 (独立 venv 运行)
  - tests/test_layout_detector.py: 布局检测器完整测试 (35 个用例)

- 配置更新:
  - .env.example: 更新布局检测配置 (新增 LAYOUT_IMGSZ, LAYOUT_DEVICE, LAYOUT_DEVICE_ID)
  - app/config.py: Settings 类对应字段
  - pyproject.toml: 新增 export 依赖组 (torch, doclayout-yolo, onnx 等)

- 删除旧文件:
  - scripts/export_picodet_onnx.py: 旧 PicoDet 导出脚本

- 文档更新:
  - README.md: 更新环境变量说明
  - 相关服务注释更新 (pdf_image_extractor.py, summary_persister.py, reextract_images.py)

此重构遵循项目初期开发阶段规范,大胆调整数据模型,无需向后兼容。
2026-06-14 10:41:44 +08:00
Rain-Bus 21f16e6756 feat: refactor summarizer and PDF extraction pipeline
- Split summarizer into summary_generator and summary_persister modules
- Refactor pdf_image_extractor to two-phase pipeline with PicoDet layout detection
- Add layout_detector service for PicoDet-S_layout_3cls integration
- Add exceptions module with ConflictError and NotFoundError
- Improve admin dashboard with better statistics and task management
- Add design review document with system optimization suggestions
- Add new tests for crawler, pdf_downloader, pipeline, and summary_utils
- Update dependencies and configuration
- Clean up dead code and improve error handling
2026-06-13 13:16:47 +08:00
Rain-Bus e2f0e1a8be feat: add claude backend, refactor summary utilities, improve batch worker pattern, add pymupdf4llm 2026-06-12 22:25:57 +08:00
Rain-Bus 18f44ac244 feat: improve PDF image extraction with caption-based labeling and fallback matching
- Enhance pdf_image_extractor with caption text extraction near images/tables
- Add figure/table type correction based on caption content
- Implement sequential numbering fallback for unmatched items
- Improve figure linking in pages with manifest ID matching and fallback strategies
- Remove docling dependency, add dev dependency group
2026-06-09 14:07:21 +08:00
Rain-Bus 32978b3fc5 feat: add admin dashboard, pipeline service, lightbox, and update dependencies 2026-06-09 09:32:10 +08:00
Rain-Bus 0d293422ac feat: enhance UI, refactor services, improve templates and tests
- Replace image_extractor with pdf_image_extractor service
- Enhance pi_client with expanded API capabilities
- Improve summarizer service with additional features
- Update admin routes with more endpoints
- Add login page template
- Enhance detail page with comprehensive layout
- Improve search and trends pages
- Update base template with additional elements
- Refactor tests for better coverage
- Add validate_summary script
- Update project configuration and dependencies
2026-06-07 19:38:58 +08:00
Rain-Bus 4a72c35452 chore: update .env.example defaults and add uv.lock 2026-06-06 01:02:47 +08:00