How to convert a scanned PDF or a photo of a table to Excel
Last updated: 17 October 2026
Scans have no text inside, so they must be read with OCR. Here is how to get the best result.
Why a scan is different
A scanned PDF or a phone photo is a picture. There is no text to copy, so the words have to be recognised first. This is called OCR (optical character recognition). The converter does it automatically when it sees that a page is a picture. Choose the language of the document under Language of scanned pages - 45 languages are available, and English is always read as well.
Getting a good scan
- Lay the page flat and photograph it straight from above, in good light, without shadows.
- Use a resolution of 200 to 300 dpi when scanning. Very small or blurry text cannot be read reliably.
- Make sure the whole table is inside the picture, including the heading row.
- A page photographed sideways is turned automatically, and a slightly crooked page is straightened.
Checking OCR results
OCR can confuse similar characters, for example the letter O with zero or l with one. Inside number columns, the "Letters mistaken for digits" fix corrects these automatically and lists every change.
Cells the OCR was unsure about are marked yellow. Compare them with the original before using the numbers. A heading too blurry to read is named "Column 2" and so on, rather than filled with a guess.