OCR, in one picture
A scanner makes a photo of the page. To the computer, that's dots. OCR finds the shapes of letters in those dots and writes the matching characters. The result is either a text file or a PDF with an invisible text layer under the image.
| Term | Meaning |
|---|---|
| Scanned PDF | Only the picture of each page. You can't select words. |
| OCR'd (searchable) PDF | Picture plus hidden text. Search and copy work. |
| OCR scanner | A scanner or app that runs OCR right after scanning. |
Free OCR on iPhone
- Live Text: select text in a photo or the camera view and copy it (Apple).
- Preview (iOS 26): Export a PDF with Embed Text on; Apple says it runs OCR and makes the PDF searchable (Apple). Steps: make a PDF searchable.
- Notes: Scan Text inserts recognized text into a note.
How accurate is it?
On sharp, printed pages, very. Accuracy drops with blur, shadows, faded receipts, unusual fonts and handwriting. Better light helps more than any setting: see how to scan clearly.
What OCR is for
Finding a word in hundreds of pages, copying a paragraph, letting screen readers read a scan, and receipt or invoice apps pulling out totals. SafeDoc uses it for search: every scan and imported PDF is read on the device, offline, in English, Portuguese and Spanish.