PDF OCR Processor

Advanced PDF processing with AI-powered OCR, text extraction, and selectable text overlays using Ollama models

🚀 Features

• AI-Powered OCR using Ollama models (llava, moondream, etc.)
• Modular Architecture with clear separation of concerns
• Multiple Output Formats:
  • SVG with selectable text overlays
  • Raw text extraction
  • JSON metadata
• Image Enhancement with multiple strategies
• Robust Error Handling with configurable retries
• Parallel Processing for batch operations
• CLI Interface with progress tracking

🛠️ System Architecture

┌─────────────────────────────────────────────────┐
│               PDF OCR Processor                 │
├─────────────────┬───────────────────────────────┤
│  ┌────────────┐ │  ┌─────────────────────────┐  │
│  │ PDF        │ │  │      OCRProcessor       │  │
│  │ Processor  ├─┼─▶│  - Text extraction      │  │
│  └────────────┘ │  │  - Ollama integration   │  │
│                 │  └─────────────┬───────────┘  │
│  ┌────────────┐ │  ┌─────────────▼───────────┐  │
│  │ Image      │ │  │      SVG Generator      │  │
│  │ Enhancer   ├─┼─▶│  - Text overlay         │  │
│  └────────────┘ │  │  - Searchable output    │  │
└─────────────────┴───────────────────────────────┘

📦 Installation

Prerequisites

• Python 3.8+
• Ollama (for OCR processing)
• System dependencies:

  # Ubuntu/Debian
  sudo apt-get install -y tesseract-ocr poppler-utils
  
  # macOS
  brew install tesseract poppler
  

Install from source

# Clone the repository
git clone https://github.com/wronai/ocr.git
cd ocr

# Create and activate a virtual environment
python -m venv venv
source venv/bin/activate  # Linux/macOS

# Install dependencies
pip install -r requirements.txt
pip install -r requirements-dev.txt  # For development

🏁 Quick Start

Basic Usage

# Process a single PDF
python -m pdf_processor --input document.pdf --output output/

# Process all PDFs in a directory
python -m pdf_processor --input ./documents --output ./output --model llava:7b

# Show help
python -m pdf_processor --help

Python API

from pdf_processor import PDFProcessor
from pdf_processor.processing.pdf_processor import PDFProcessorConfig

# Configure the processor
config = PDFProcessorConfig(
    input_path="document.pdf",
    output_dir="./output",
    ocr_model="llava:7b",
    dpi=300,
    max_workers=4
)

# Process a document
processor = PDFProcessor(config)
result = processor.process_pdf("document.pdf")
print(f"Processed {result['pages_processed']} pages")

⚙️ Configuration

Configuration File

Create a config.yaml file:

# config.yaml
input_path: ./documents    # Input file or directory
output_dir: ./output       # Output directory
ocr_model: llava:7b        # Ollama model to use
dpi: 300                   # Image resolution
max_workers: 4             # Number of worker threads
timeout: 300               # Timeout in seconds
max_retries: 3             # Max retry attempts
log_level: INFO            # Logging level
log_file: pdf_processor.log # Log file path

# Image enhancement strategies
enhancement_strategies:
  - original            # Keep original image
  - grayscale           # Convert to grayscale
  - adaptive_threshold  # Apply adaptive thresholding
  - contrast_stretch    # Stretch contrast
  - sharpen             # Sharpen image
  - denoise             # Remove noise

Environment Variables

export OLLAMA_HOST="http://localhost:11434"
export OLLAMA_MODEL="llava:7b"
export LOG_LEVEL="DEBUG"

🚀 Advanced Usage

Processing Options

# Process with specific DPI
python -m pdf_processor --input document.pdf --output output/ --dpi 400

# Limit number of pages to process
python -m pdf_processor --input document.pdf --output output/ --max-pages 10

# Use a specific enhancement strategy
python -m pdf_processor --input document.pdf --output output/ --enhance grayscale

# Process in verbose mode
python -m pdf_processor --input document.pdf --output output/ --verbose

Available Enhancement Strategies

• original: Keep original image (fastest)
• grayscale: Convert to grayscale (good for text-heavy documents)
• adaptive_threshold: Apply adaptive thresholding (good for low-quality scans)
• contrast_stretch: Stretch contrast to improve readability
• sharpen: Apply sharpening filter
• denoise: Remove image noise

🛠️ Development

Project Structure

pdf_processor/
├── __init__.py          # Package initialization
├── cli.py               # Command-line interface
├── config/              # Configuration files
├── models/              # Data models
│   ├── __init__.py
│   ├── ocr_result.py    # OCR result data structures
│   └── retry_config.py  # Retry configuration
├── processing/          # Core processing modules
│   ├── __init__.py
│   ├── image_enhancement.py  # Image processing
│   ├── ocr_processor.py      # OCR processing
│   ├── pdf_processor.py      # Main PDF processing
│   └── svg_generator.py      # SVG output generation
└── utils/               # Utility functions
    ├── file_utils.py    # File operations
    ├── logging_utils.py # Logging configuration
    └── validation_utils.py # Input validation

Running Tests

# Install test dependencies
pip install -r requirements-dev.txt

# Run all tests
pytest

# Run tests with coverage report
pytest --cov=pdf_processor --cov-report=html

🤝 Contributing

Contributions are welcome! Please follow these steps:

1. Fork the repository
2. Create a feature branch (git checkout -b feature/amazing-feature)
3. Commit your changes (git commit -m 'Add some amazing feature')
4. Push to the branch (git push origin feature/amazing-feature)
5. Open a Pull Request

📄 License

This project is licensed under the Apache License 2.0 - see the LICENSE file for details.

📚 Resources

• Ollama Documentation
• PyMuPDF Documentation
• Pillow Documentation

🙏 Acknowledgments

• The Ollama team for their amazing AI models
• The PyMuPDF team for excellent PDF processing
• All contributors who have helped improve this project

🛠️ Development Workflow

This project uses a script-based workflow for development tasks. All scripts are located in the scripts/ directory and can be run directly or via the Makefile.

Setup

1. Clone the repository and navigate to the project directory:

   git clone https://github.com/wronai/ocr.git
   cd ocr
   

2. Set up the development environment:

   make install-dev
   

   This will:

   • Create and activate a virtual environment
   • Install all development dependencies
   • Set up pre-commit hooks

Common Development Tasks

# Run tests
make test

# Run tests with coverage
make test-cov

# Format code
make format

# Run linters
make lint

# Start development server
make dev-server

# Build documentation
make docs
make docs-serve  # Serve docs locally

Scripts Directory

All development and build scripts are located in the scripts/ directory. See scripts/README.md for detailed documentation of each script.

Docker Development

# Build Docker image
make docker-build

# Start services with Docker Compose
make docker-run

# Stop services
make docker-stop

🤝 Contributing

Contributions are welcome! Please follow these steps:

1. Fork the repository
2. Create a feature branch (git checkout -b feature/amazing-feature)
3. Commit your changes (git commit -m 'Add some amazing feature')
4. Push to the branch (git push origin feature/amazing-feature)
5. Open a Pull Request

Please ensure your code follows our coding standards and includes appropriate tests.

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

📜 Changelog

See CHANGELOG.md for a list of changes in each version.**

python proc.py --model llava:7b --workers 4

3. View Results
   • Open output/*_complete.svg in your browser
   • Check details in output/processing_report.json

📚 Documentation

Full documentation is available in the docs/ directory:

• 📖 User Guide
• ⚙️ Installation and Configuration
• 🔧 API Reference
• ❓ Frequently Asked Questions
• 👨‍💻 Development and Contributing