# Edit a scanned PDF after OCR: a worked example

> Change B-108 to B-109 on a fictional scan. Follow the OCR and text-box settings, then inspect the image-only PDF export.

[Read this guide in Edit PDF Pages](https://editpdfpages.com/guides/edit-scanned-pdf/)

Content updated: 2026-10-12. Example checked 12 October 2026 in Chrome 155 on macOS.

This scanned note needs a corrected reference. OCR finds B-108 so you can select it; the text editor then places B-109 over the old wording. Recognition and replacement are two separate steps.

Use the fictional [scanned note](https://editpdfpages.com/guides/examples/scanned-note.pdf) in the [PDF workspace](https://editpdfpages.com/). The visible correction requires sanitized image-only export. If you only need to find words in the unchanged scan, use the [searchable-PDF guide](https://editpdfpages.com/guides/make-scanned-pdf-searchable/).

## Recognize and select the reference

1. Choose **Recognize text**, leave **Pages** on **Current page**, and choose **Recognize text** inside the dialog.
2. When **Text is ready** appears, choose **Edit recognized text**.
3. Click or tap **B-108**. Compare the selected wording with the image, then type **B-109** directly on the page. The sidebar's **Replacement text** stays synchronized.

Leave **Select nearby paragraph lines** off for this small change. English OCR can confuse similar letters and digits, so the selected text needs checking even when it looks plausible. If a word was missed, try another recognized area or deliberately re-run OCR; a poor scan may return the same mistake.

## Give the correction enough room

The checked example used these settings:

| Setting | Value for this note |
| --- | --- |
| Text | B-109 |
| Font | Sans-serif, 23 pt, bold |
| Box width (% page) | 20 |
| Box height (% page) | 5 |

These dimensions give the replacement room at that font size. They are a starting point for this note, not settings to apply to every scan. Exact font recovery from the image is unavailable; compare the result with the surrounding line.

New replacements start with **Auto-expand box**. Entering dimensions or dragging the lower-right corner switches to **Keep size**. If the text overflows, enlarge the box, return to Auto-expand box or choose **Shrink to fit** under **Text fit**. Export remains blocked until it fits.

Check nearby words and borders after resizing. Surrounding content stays in place, and a larger box can overlap it. Choose **Done** above the page when you are ready to review.

## Check the exported B-109

Choose **Export** and the **PDF** format. The mode is **Sanitized image-only PDF**. Select a page image quality and choose **Create PDF**.

The exported pages are rebuilt without source PDF objects or OCR layers. The corrected copy therefore loses searchable text, interactive forms and links. Reopen it and inspect B-109 at normal reading size, then zoom in on the edges of the correction and the covered old reference.

Keep a **.folio** project only as an editable backup: it contains the original scan and OCR results. If the finished page is needed as an image instead, [export the edited page as PNG](https://editpdfpages.com/guides/pdf-to-images/).

## Worked example

![OCR-recognized reference changed to B-109, with a 23 point bold text box and enlarged dimensions](https://editpdfpages.com/guides/images/edit-scanned-pdf.png)

OCR found B-108, then the visual replacement changed it to B-109. The example uses a 23 pt bold replacement in a larger box. The sanitized download has no searchable OCR layer.

## Related guides

- [How to make a scanned PDF searchable](https://editpdfpages.com/guides/make-scanned-pdf-searchable/)
- [Replace text in a PDF: what changes when you export](https://editpdfpages.com/guides/replace-text-in-pdf/)
- [How to convert PDF pages to JPG or PNG](https://editpdfpages.com/guides/pdf-to-images/)

These guides were prepared with AI assistance and checked using fictional documents. Examples contain no real customer or account data. See each guide's example for the workflow and its limits.