OCR PDF

How to OCR a scanned PDF

1. Select the scanned PDF or drop it onto the page.
2. Choose the document's language and the pages to recognise.
3. Click start. Each page is analysed in your browser; long documents take a little while.
4. Download the searchable PDF, or the recognised text.

What OCR does

A scanned PDF is a set of photographs of paper. You can read it, but your computer can't: searching, copying and screen readers find nothing. Optical character recognition analyses the shapes in each page image and works out the words. The searchable PDF keeps your original scan as it looks and adds an invisible text layer aligned with the words, so you can search, select and copy text while the page appears unchanged.

Getting accurate results

• Pick the right language. Recognition uses language-specific models, and the wrong one causes many errors.
• Scan at 300 DPI if you can. Low-resolution, blurry or skewed scans produce more mistakes.
• Check important details such as names, figures and dates. OCR is very good on clean printed text but it is never perfect, and handwriting is generally not recognised.

Private OCR in your browser

Recognition uses Tesseract, an open-source OCR engine, running in your browser. The first time you use a language, its recognition data is downloaded once from a public server; your document itself is never uploaded. Each run recognises up to 50 pages, so long scans can be done in parts. After OCR, you can extract the text with PDF to Text or convert to Word.

Frequently asked questions