feat: v2.0.0 - PPU model support, 100+ languages, and comprehensive d…

…ocumentation

- Added PPU PaddleOCR model support with specialized preprocessing
- Fixed PPU model recognition issues (red channel grayscale, 0-based indexing)
- Prevented stack overflow errors on large documents
- Extended language support from 14 to 100+ languages
- Added comprehensive documentation (usage, API, troubleshooting, models)
- Added screenshots to README
- Improved error handling and memory management
- Updated npm package configuration for v2.0.0 release

BREAKING CHANGES: Major version update with significant API enhancements

feat: Add table detection, layout analysis, and unified model registr…

…y (v1.3.0)

- Integrated RapidTable for table structure recognition
  - PP-Structure models for English and Chinese
  - SLANet+ model for enhanced accuracy
  - HTML table output with cell detection

- Integrated RapidLayout for document layout detection
  - PP Layout CDLA model
  - YOLOv8n Layout model for academic papers
  - DocLayout-YOLO for DocStructBench
  - Detects text, titles, tables, figures, and formulas

- Created unified model registry
  - Local models from OnnxOCR directory
  - Local models from ppu-paddle-ocr directory
  - Remote models from RapidOCR
  - PPU Paddle OCR English mobile set as default

- Enhanced UI with:
  - Processing mode selector (OCR, Table, Layout, All-in-One)
  - Model Manager tab with GitHub links to sources
  - Model source information display
  - Configurable model defaults

- Updated PP-OCRv5 model URLs to use master branch
- Added camera, clipboard, and history features
- Created workers for table and layout detection

🤖 Generated with [Claude Code](https://claude.ai/code)

Co-Authored-By: Claude <[email protected]>