How to convert a scanned PDF to Word

Why a scanned PDF needs OCR first

A scanner or a phone camera takes a picture of the page. To a computer there are no words in that picture, only dots of colour. A normal PDF to Word conversion looks for text, finds none, and gives you a document with pictures of the pages, or an empty one.

Text recognition (OCR) reads the letters in the picture and turns them into real text. A quick test: try to select a word in the PDF. If nothing is selected, or the whole page is selected as one picture, it's a scan.

How to convert a scanned PDF to Word in PDFZyro

The PDF itself isn't changed. The Word file uses the recognized text and leaves out the picture of each scanned page, so you get text you can edit rather than a picture with text on top.

1. 1Open PDF to Word on the Home page and choose the scanned PDF from your documents, or upload it.
2. 2Choose All pages, or Some pages and type them, for example 1-3, 5, 8-.
3. 3Tick Recognize text in scanned pages (OCR).
4. 4Choose the document's language, or up to three if it mixes languages. English, Spanish, French, German, Italian, Portuguese, Dutch and Russian are available.
5. 5Click Convert to Word. Recognition and conversion run in the background; you can follow the progress and stop it if you need to.
6. 6When it's ready, the download starts. The notes say how many scanned pages were recognized and list any pages that were hard to read.
7. 7Open the .docx in Word, Google Docs, LibreOffice or Pages and check it against the scan.

Or make the scan searchable first

If you'll keep using the PDF as well, run Make searchable (OCR) on it first, then convert it with PDF to Word without ticking the OCR box. You get a searchable PDF, saved as a new version with the original scan kept, and a Word file made from it.

This is also the way to go when you want to search, copy or redact the scan in PDFZyro. On the Free plan it uses two daily tasks, one for OCR and one for the conversion, instead of one.

How to get accurate text from a scan

• Scan at 300 DPI. At lower resolutions small letters blur and OCR makes more mistakes.
• Keep pages straight and flat, with dark text on a light background. Avoid shadows and angles when you photograph a page with a phone.
• Choose the right language. OCR uses each language's letters and words, so the wrong one gives more mistakes and can lose accents.
• Check numbers, names and email addresses closely. Similar characters, such as 0 and O or 1 and l, are the most common mistakes.
• Look first at the pages listed as hard to read. Printed and typed text is recognized well; handwriting usually isn't.

What to expect from the Word file

• Paragraphs, headings and lists come through as real Word formatting, and tables become Word tables.
• The text is set in a standard font, so it won't match the typeface of the scan exactly.
• Simple layouts convert well. Columns, forms and text boxes may need some touching up.
• Anything that isn't recognized as text, such as a handwritten note, a signature or a logo in the scan, isn't in the Word file. Keep the PDF as the record of the original.

Other ways to convert a scanned PDF to Word

• Microsoft Word: choose File, then Open, and pick the PDF. Word converts it into an editable document. It works best with PDFs that already contain text; with scans the result varies, so check it carefully.
• Google Drive: upload the PDF, right-click it and choose Open with, then Google Docs. The text is recognized into a new Doc; choose File, then Download, then Microsoft Word (.docx). The layout and pictures are usually lost.
• Adobe Acrobat: the paid version can export a scanned PDF to Word with text recognition. The free Acrobat Reader can't.
• Retyping: for a page or two of handwriting, typing it out is often quicker than correcting OCR.