Tesseract

A Model Context Protocol (MCP) server that provides OCR capabilities for PDF documents using Tesseract OCR. This server allows AI assistants to extract text from PDF files, supporting multiple languages including English and Simplified Chinese out of the box.

Features

• PDF to Text Conversion: Extract text from PDF documents using OCR technology
• Multi-language Support: Process documents in multiple languages (English and Simplified Chinese by default)
• Dockerized Solution: Easy deployment with Docker
• MCP Integration: Seamlessly integrates with AI assistants that support the Model Context Protocol

Prerequisites

• Docker installed on your system

Build Instructions

Build the Docker image with the following command:

docker build -t tesseract-pdf-mcp .

Running the Server

Run the MCP server with the following command:

docker run -it --rm \
  -v /path/to/your/pdfs:/pdfs \
  tesseract-pdf-mcp

Important Notes:

• The -v /path/to/your/pdfs:/pdfs option mounts a volume from your host system to the Docker container, allowing the server to access PDF files.
• Replace /path/to/your/pdfs with the actual path to the directory containing your PDF files.
• The server will be accessible via standard input/output (stdio) as specified in the MCP protocol.

Usage

The server provides a tool called convert_pdf that can be used to extract text from PDF files.

Input

The convert_pdf tool accepts the following JSON input:

{
  "file_path": "/pdfs/document.pdf",
  "language": "eng"
}

Parameters:

• file_path (required): Path to the PDF file to process. This should be the path inside the container (e.g., /pdfs/document.pdf).
• language (optional): Language for OCR processing. Default is "eng" (English).
  • Available languages by default: "eng" (English), "chi_sim" (Simplified Chinese)

Output

The tool returns a JSON response with the following structure:

{
  "status": "success",
  "output_path": "/pdfs/document.txt"
}

On success:

• status: Will be "success"
• output_path: The absolute path to the generated text file

On error:

• status: Will be "error"
• message: Error description
• output_path: Will be null

Example Usage

When connected to an AI assistant that supports MCP:

1. The assistant can use the convert_pdf tool to extract text from a PDF file
2. The text file will be created in the same directory as the PDF file
3. The assistant can then access the text file to analyze its contents

Connecting to AI Tools

To connect this MCP server to AI tools that support the Model Context Protocol, you'll need to configure the tool with the appropriate settings.

Configuration Example

Add the following configuration to your AI tool's settings:

{
  "mcpServers": {
    "tesseract-pdf-mcp": {
      "command": "docker",
      "args": [
        "run",
        "-i",
        "--rm",
        "-v",
        "/path/to/your/pdfs:/pdfs",
        "tesseract-pdf-mcp"
      ],
      "disabled": false,
      "autoApprove": []
    }
  }
}

Make sure to replace /path/to/your/pdfs with the actual path to your PDF files directory.

Usage with AI Tools

Once connected:

1. The AI tool will have access to the convert_pdf tool provided by this MCP server
2. You can ask the AI to extract text from PDF documents
3. The AI will use the MCP server to process the PDFs and access the resulting text

Adding More Languages

The server comes with English (eng) and Simplified Chinese (chi_sim) language support by default. To add more languages:

1. Modify the Dockerfile by adding additional language packs to the apt-get install command:

RUN apt-get update && apt-get install -y --no-install-recommends \
    tesseract-ocr \
    tesseract-ocr-eng \
    tesseract-ocr-chi-sim \
    tesseract-ocr-fra \    # Add French
    tesseract-ocr-deu \    # Add German
    tesseract-ocr-spa \    # Add Spanish
    poppler-utils \
    && apt-get clean \
    && rm -rf /var/lib/apt/lists/*

2. Rebuild the Docker image:

docker build -t tesseract-pdf-mcp .

Available Language Codes

Some common language codes for Tesseract OCR:

• eng: English
• chi_sim: Simplified Chinese
• chi_tra: Traditional Chinese
• fra: French
• deu: German
• spa: Spanish
• ita: Italian
• jpn: Japanese
• kor: Korean
• rus: Russian

For a complete list of available language packs, refer to the Tesseract documentation.

Debugging Inside the Container

If you need to debug or test the PDF conversion logic directly inside the container, follow these steps:

Starting an Interactive Shell

Launch an interactive shell session in the container with the following command:

docker run --rm -it -v /path/to/your/pdfs:/data tesseract-pdf-mcp /bin/bash

This command:

• Creates a container from the tesseract-pdf-mcp image
• Mounts your local PDF directory to /data inside the container
• Overrides the default command to start a bash shell
• Removes the container automatically when you exit (--rm)

Working Inside the Container

Once inside the container's shell, you can:

• Navigate the filesystem using standard Linux commands (cd, ls, etc.)
• Access your mounted PDFs in the /data directory
• Run Python scripts or start an interactive Python session

Testing the Conversion Function

You can test the PDF to text conversion directly using Python's interactive shell:

# Start Python interactive shell
python3

# Import the conversion function
from ocr.converter import pdf_to_text

# Process a PDF file (replace with your actual filename)
output_path = pdf_to_text('/data/my_document.pdf', lang='eng')

# Verify the result
print(f"Conversion successful. Output saved to: {output_path}")

# Exit Python shell
exit()

The converted text file will be saved in the same directory as your PDF file (in the /data directory), making it accessible from your host machine as well.