How to OCR a PDF and Make It Searchable
Last updated · By the PDFSolution team
A scanned page is a photograph. Your computer sees pixels, not words, which is why searching returns nothing. Optical character recognition analyses those pixels and works out which characters they represent.
Recognition quality is dominated by input quality. There is far more to gain from a better scan than from any setting.
Ready to do it now? Open the tool this guide describes.
OCR a PDFStep by step
- 1
Start with the best scan you have
300 DPI, straight, high contrast, greyscale or black and white. Re-scan if the original is faint.
- 2
Upload the PDF and choose the language
The language setting tells the recogniser which characters and word shapes to expect, and matters a lot.
- 3
Run recognition
Every page is analysed on your own device, so long documents take a few minutes.
- 4
Choose your output
Take the plain text, a searchable PDF that looks unchanged, or an editable Word document.
Useful tips
- Straighten crooked scans before running OCR; skew is one of the biggest causes of errors.
- Always proofread numbers. Digits are where misreads cause real damage.
- Handwriting is not supported — OCR is designed for printed type.
Common problems
- The results are full of nonsense characters
- Usually the wrong language is selected, or the scan is too low resolution. Fix those two things first.
- OCR is very slow
- Recognition is computation-heavy and runs locally. Close other tabs, or process the document in smaller parts.
- Nothing was recognised at all
- Check that the pages are not blank or extremely faint when zoomed in.
Frequently asked questions
- What accuracy should I expect?
- A clean 300 DPI printed page usually reads very well. Faxes, photocopies of photocopies and photographs of screens read much worse.
- Does OCR change how the page looks?
- Not when you create a searchable PDF: the text layer sits invisibly behind the original image.
- Is my scan uploaded?
- No, recognition runs in your browser.
Related guides
- How to Convert a PDF to WordWhat converts cleanly, what does not, and how to handle scanned documents that contain no text at all.
- How to Convert a Scanned PDF to WordRecover the wording from paper documents without retyping — and know where to proofread.
- How to Translate a PDF Into Another LanguageTranslate a document you cannot read — including scans — and know when a human translator is still required.