You have received something that looks like a PDF but behaves like a photograph: nothing can be selected and nothing corrected. Here is the full route to turning it back into an editable document.
▶ 1. Understanding what you have in front of you
A scanned PDF is not a faulty PDF: it is simply a container that holds the image of a page instead of text.
The test is immediate: try selecting a word. If nothing highlights and you find yourself drawing a rectangle, the document is a scan. That is why no editor manages to “correct the typo”: technically, that typo does not exist as text.
The solution therefore starts with a recognition step. With 123 PDF Editor PRO the whole route runs in the browser tab, without installing desktop software and without uploading the document — a relevant point, since scans are often contracts, medical reports or identity documents.
🔒
Everything happens on your device: the document is never uploaded to any server and the editor works even without a connection.
▦ 2. Preparing the pages
Before recognising the text it is worth sorting out the structure: recognition run on crooked or duplicated pages gives mediocre results.
Straightening and reordering the pages before OCR noticeably improves the result.
Open the file and go into Organize pages.
Straighten crooked pages with Rotate left or Rotate right.
Delete blank pages and duplicates produced by the scanner.
If the margins are huge, trim them with Crop page.
Confirm with Apply page changes.
◎ 3. Recognising the text
This is the step that turns the image into a proper document.
If the Scanned pages detected window appears, carry on from there; otherwise open Local OCR.
Select the language of the document, or leave Automatic if you are not sure.
Click Run local OCR and wait for processing to finish.
Check on one page that the text is now selectable.
After recognition the flat scan turns into editable, searchable text.
✍️ 4. Correcting the text
After recognition the document is edited like any other PDF, with one caveat: recognised text may contain the odd reading error.
With the block selected, you work on text, font, size and colour.
Choose Word mode for pinpoint corrections or Text block to rewrite whole sentences.
Click the text and correct it; the panel shows the font, size and colour detected.
Pay particular attention to numbers, dates, codes and proper names, where recognition errors are most frequent.
Use Search and replace if the same error repeats throughout the document.
💡
On a scan, recognised text does not always carry the same font as the printed original. If a correction looks out of place, match font and size manually using a nearby word as your reference.
▧ 5. Images, stamps and signatures
Scans often contain graphic elements that need handling separately from the text.
Switch to ▣ Select image mode to select logos, stamps or boxes.
Replace an outdated logo with Replace with local image (PNG or JPG).
Delete elements that are no longer relevant with Remove image.
To add a signature, use Add signature: the procedure is described in the guide on how to sign a PDF.
⤓ 6. Saving and archiving
A scanned and recognised document is worth far more than the original, because it becomes searchable: it deserves to be archived properly.
Click Download PDF to save the edited version.
Scans are often heavy: if you have to email them, compress them following the guide on how to compress a PDF, keeping the option that preserves selectable text switched on.
If the document contains personal data, consider Password protection.
Use Save project if the job is long and you want to come back to it later.
? 7. Frequently asked questions
The questions that come up most often on this topic.
Can I edit a scan without running OCR?
You can work on the structure — rotate, crop, delete pages — and lay new elements on top with Add text. But to correct existing text, recognition is essential: without it, that text does not exist.
Does the document stay identical to the original after OCR?
The look of the page stays as it was, with the original image in place; recognition adds the text on top, making it selectable and editable.
Is handwriting recognised?
No: OCR reads printed characters. On handwritten notes, signatures and pen annotations it produces no useful result.
Are my scans uploaded anywhere?
Never. Recognition, editing and saving all happen in the browser tab, on your device.
Bring your scanned document back to life
Install the extension, open the scan and make it editable — with nothing uploaded online.