Offline Chinese OCR with Pinyin

A simple Chinese OCR web app that runs entirely in your browser.

Convert images containing Chinese text into editable text and automatically display Pinyin pronunciation.

🌐 Web App

Try it online:

https://iwaoyuiko.github.io/web-ocr-app/

🔒 Images never leave your device. All processing happens locally in your browser.

Features

• 📷 OCR for Chinese text from images
• 🔤 Automatic Pinyin display
• 🔒 Fully offline — no image uploads
• ⚡ Runs directly in the browser
• 📋 Copy recognized text with one click
• 🆓 Free to use

How to Use

1. Select an image (JPG or PNG)
2. Click Run OCR
3. View the recognized Chinese text
4. Check the Pinyin pronunciation below
5. Copy the text if needed

Privacy

All processing happens locally in your browser.

Your images are never uploaded to any server.

Tech Stack

OCR

• PP-OCRv5
• ONNX Runtime Web
• ppu-paddle-ocr

Chinese Processing

• pinyin-pro

Frontend

• React
• TypeScript
• Vite

License

This project is released under the MIT License.

Third-Party Components

This project uses PP-OCRv5 models derived from PaddleOCR.

PaddleOCR is licensed under the Apache License 2.0.

Acknowledgements

This project is built with the help of the following open-source projects:

• PaddleOCR
• ONNX Runtime Web
• ppu-paddle-ocr
• pinyin-pro

Many thanks to the authors and contributors of these projects.