How to copy text from a scanned PDF (OCR)
You click on a scanned contract to copy one sentence – and nothing gets selected. That’s not a bug: a scan is a picture of a page, not text. Text recognition (OCR) turns it back into letters.
Why you can’t select text in a scan
A PDF can hold two kinds of content. A document saved from Word or an accounting system has a text layer: letters are stored as characters, so you can select, copy and search them. A scan or a photo of a document is an image – to a computer it’s just coloured dots. To get text out of it, it has to be recognised: that’s OCR (optical character recognition).
A quick test: press Ctrl+F in your PDF viewer and type a word you can see on the page. If nothing is found, the file has no text layer.
How to extract text from a scan in PDFowo
- Open the PDF to Text tool and add your PDF. You can also upload just a photo of the document (JPG, PNG, WEBP).
- Choose the language of the document – this decides whether accented letters are recognised correctly.
- Leave recognition on “Automatic”: OCR runs only on pages without text, and the other pages are read straight from the file.
- Click “Extract text”. Copy the result with one button or download it as a .txt file.
The first time, your browser downloads the language data OCR needs, so it takes a little longer to start. The document itself is never uploaded – recognition happens on your computer or phone.
How to get better recognition
- A straight page: a skewed scan means more mistakes. Take the photo from directly above or use the document scanner, which straightens and crops the page.
- Good contrast: dark text on a light background works best. Avoid shadows, lamp glare and creases.
- Enough resolution: small print needs a sharp image – on a scanner, choose about 300 dpi.
- “Whole document” mode: if text taken from the PDF comes out garbled or in the wrong order, treat every page as a scan.
What OCR can’t do
Recognition handles printed text well, but handwriting usually comes out poorly. Tables end up as lines of text rather than cells. Always check the result, especially numbers: amounts, account numbers and dates – one wrong digit can change the meaning of a document.
If you’d rather fix the wording in the document itself, open it in the PDF editor – there’s an OCR button above the page there too.
Free, no sign-up. Your files stay on your device.
Open the PDF to Text tool →Example: quoting a scanned contract
Your bank sent a scanned loan agreement and you want to paste the interest-rate clause into an e-mail. Instead of retyping it, you upload the PDF to PDF to Text, pick English and copy the paragraph a few seconds later. Before sending, you compare the numbers with the original – OCR may have confused “0” with “O”.
Frequently asked questions
Is OCR free?
Yes. Text recognition in PDFowo is free and needs no account.
Is my document uploaded to a server?
No. Your browser only downloads the language data OCR needs; the file stays on your device.
Why are there mistakes in the text?
Usually because of a poor scan: a skewed page, shadows or low resolution. Take a sharper photo and check that the right language is selected.