Image OCR
Extract text from images directly inside VS Code — right-click any image in the Explorer and get OCR results in a new editor tab instantly.
Built by Manoj Patode using Tesseract.js.
🔒 100% local & private — all OCR processing happens on your machine. No API calls, no data ever leaves your computer.
Features
- 🖱️ Right-click any image in the Explorer →
Extract Text (OCR)
- 📄 Opens result in a new editor tab — ready to copy, edit, or save
- 📝 Single-line output — all whitespace collapsed for clean results
- 🖼️ Supports all common formats:
| Format |
Extension |
| PNG |
.png |
| JPEG |
.jpg, .jpeg |
| WebP |
.webp |
| BMP |
.bmp |
| TIFF |
.tiff, .tif |
| GIF |
.gif |
| SVG |
.svg (auto-converted to PNG first) |
Usage
- Open any project folder in VS Code
- Right-click an image file in the Explorer sidebar
- Click Extract Text (OCR)
- A progress notification appears while OCR runs
- The extracted text opens in a new plaintext editor tab
First run: Dependencies (tesseract.js + sharp) are automatically installed on first use. This takes 15–30 seconds. Subsequent runs are instant.
Requirements
- VS Code
^1.85.0
- Node.js and npm must be available on your
PATH (for first-run dependency install)
- Internet connection on first use (to download Tesseract.js and Sharp)
Extension Settings
This extension has no configurable settings currently. Future versions will support:
- Language selection for OCR (default: English)
- Multi-line vs single-line output toggle
- Copy to clipboard option
Known Issues
- First-run delay: Dependency install takes 15–30s on first use — this is expected.
- SVG with external assets: SVGs that reference external fonts or images may not render perfectly during conversion.
- Handwritten text: OCR accuracy is best on printed/digital text. Handwriting recognition is limited.
- Low-resolution images: Very small or blurry images may produce poor results.
Release Notes
0.0.2
- Added clipboard OCR via status bar button (
OCR Clipboard)
- Fixed clipboard read to use PowerShell instead of Electron (was failing with "Cannot read properties of undefined")
- Extension now activates on VS Code startup
0.0.1
- Initial release
- OCR support for PNG, JPG, JPEG, WEBP, BMP, TIFF, GIF, SVG
- Single-line formatted output
- Auto dependency installation on first use
About
Made with ❤️ by Manoj Patode
- Powered by Tesseract.js — pure JavaScript OCR engine
- SVG conversion via Sharp
- Runs entirely offline/local — no external API, your images and text never leave your machine
| |
