ocr-mcp
OCR MCP 服务
基于 MCP 的 OCR 服务,由 PaddleOCR 驱动,支持文本识别和表格识别。
功能特性
OCR 文本识别:识别图像中的文本内容,支持中文、英文等多种语言
表格识别:识别图像中的表格并返回结构化数据
手写识别:识别图像中的手写文本
公式识别:识别图像中的数学公式并返回 LaTeX 格式
多格式支持:支持 JPG、PNG、BMP、TIFF、WebP 等图像格式
URL 支持:支持本地文件路径和网络 URL
MCP 协议:标准 MCP 服务,可集成到兼容 MCP 的 AI 助手中
Related MCP server: mcp_ocr
安装
通过 pip 安装
pip install ocr-mcp从源码安装
git clone https://github.com/LinuxLinking/OCR_MCP.git
cd OCR_MCP
pip install -e .使用
作为 MCP 服务运行
# Run directly
ocr-mcp
# Or run via Python module
python -m ocr_mcp与 Claude Desktop 集成
添加到 Claude Desktop 配置文件:
{
"mcpServers": {
"ocr": {
"command": "ocr-mcp",
"args": []
}
}
}MCP 工具说明
ocr_recognize
识别图像中的文本内容。
参数:
source(必填):图像来源,本地文件路径或 URLlanguage(可选):识别语言,默认为ch(中文),支持en、japan、korean等use_angle_cls(可选):是否启用文本方向分类,默认为true
示例输出:
1. Hello World (confidence: 98.50%)
2. 你好世界 (confidence: 97.20%)ocr_table
识别图像中的表格内容并返回结构化数据。
参数:
source(必填):图像来源,本地文件路径或 URLlanguage(可选):识别语言,默认为ch
示例输出:
+--------+-----+------+
| Name | Age | City |
+--------+-----+------+
| Alice | 25 | NYC |
+--------+-----+------+
| Bob | 30 | LA |
+--------+-----+------+ocr_handwrite
识别图像中的手写文本内容。
参数:
source(必填):图像来源,本地文件路径或 URLlanguage(可选):识别语言,默认为ch
示例输出:
1. Hello World (confidence: 85.50%)
2. 你好 (confidence: 82.30%)ocr_formula
识别图像中的数学公式并返回 LaTeX 格式。
参数:
source(必填):图像来源,本地文件路径或 URL
示例输出:
Formula 1:
LaTeX: E = mc^2
Formula 2:
LaTeX: \int_{0}^{\infty} e^{-x^2} dx = \frac{\sqrt{\pi}}{2}开发
安装开发依赖
pip install -e ".[dev]"运行测试
pytest tests/ -v项目结构
OCR_MCP/
├── src/ocr_mcp/
│ ├── __init__.py # Package initialization
│ ├── __main__.py # Module entry point
│ ├── server.py # MCP service main entry (dispatch layer)
│ ├── ocr_engine.py # Subprogram 1: Text recognition engine
│ ├── table_parser.py # Subprogram 2: Table recognition engine
│ ├── formula_engine.py # Subprogram 3: Formula recognition engine
│ └── utils.py # Utility functions
├── tests/
│ └── test_ocr.py # Test files
├── pyproject.toml # Project configuration
├── LICENSE # MIT License
└── README.md # Project documentation架构
本项目采用子程序并行架构,包含 4 个独立的识别工具,可单独调用:
┌─────────────────────────────────────────────────────────┐
│ MCP Server (Dispatch Layer) │
├──────────────┬──────────────┬─────────────┬─────────────┤
│ Subprogram 1 │ Subprogram 2 │ Subprogram 3│ Subprogram 4│
│ocr_recognize │ ocr_table │ocr_handwrite│ ocr_formula │
│Text Recognition│Table Recognition│Handwriting Recognition│Formula Recognition│
└──────────────┴──────────────┴─────────────┴─────────────┘
↓ ↓ ↓ ↓
OCREngine TableParser OCREngine FormulaEngine特性:
每个子程序独立运行,无依赖
支持并行调用,可同时处理多个任务
每次调用创建独立实例,避免状态冲突
可单独扩展或替换任何子程序
依赖
PaddleOCR:核心 OCR 识别引擎
PaddlePaddle:深度学习框架
OpenCV:图像处理
MCP:模型上下文协议服务框架
许可证
本项目基于 MIT 许可证 开源。
相关链接
This server cannot be installed
Maintenance
Resources
Unclaimed servers have limited discoverability.
Looking for Admin?
If you are the server author, to access and configure the admin panel.
Related MCP Servers
- FlicenseNot gradedqualityDmaintenanceProvides OCR services powered by Google's Gemini API to extract text from images via file paths or base64 strings. It enables high-accuracy text recognition and CAPTCHA processing through simple MCP tools.7
- AlicenseAqualityAmaintenanceModular OCR MCP server supporting Apple Vision, PaddleOCR, and PaddleOCR-VL backends. Enables text, layout, table, formula, and chart extraction from images via natural language.14MIT
- AlicenseNot gradedqualityBmaintenanceEnables image analysis via OpenAI-compatible vision APIs, supporting local files, URLs, and base64 inputs with intelligent tiling for high-resolution images. Provides a secure, configurable MCP stdio server for structured vision analysis.1,2641MIT
- AlicenseAqualityAmaintenanceEnables non-vision LLMs to understand images, extract text via OCR, and parse documents through a unified MCP interface, with local-first processing and optional OpenAI-compatible channels.3MIT
Related MCP Connectors
OCR, transcription, file extraction, and image generation for AI agents via MCP.
OCR.space MCP — wraps the OCR.space API (ocr.space) for image/PDF → text OCR.
ImagePay MCP: 40 x402 paid API tools (OCR, PDF, image, QR, tabular, web). Pay per call on Base.
Latest Blog Posts
- Who's Calling? MCP Hosts Are an Identity Blind Spot (And the Spec Knows It)By Om-Shree-0709 on .mcpAgent IdentityOAuth 2.1
- Your AI Chatbot Just Exposed Your CEO's Salary to an InternBy Om-Shree-0709 on .Agent IdentityMCP SecurityOAuth Delegation
- Why MCP Servers Need Execution Sandboxing (And Why Your Current Stack Isn't Enough)By Om-Shree-0709 on .Agentic AiPrompt InjectionWebAssembly
MCP directory API
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/LinuxLinking/ocr-mcp'
If you have feedback or need assistance with the MCP directory API, please join our Discord server