Digital PDF vs scanned PDF
A digital PDF (exported from Word, a website, an accounting system) contains the characters themselves, with their positions. You can select and copy the text, and any summarizer can read it.
A scanned PDF (from a scanner, a copier or a phone app) contains images. There are no characters — just pixels that look like letters. Copy-paste returns nothing, and a summarizer that only reads the text layer sees an empty document.
Some PDFs are hybrid: a scanned image with an invisible OCR text layer added by the scanner software. Those behave like digital PDFs, with the accuracy of the OCR that produced them.
How to check if your PDF is scanned
- Try to select a word. If the whole page highlights like a picture, or nothing highlights, it's scanned.
- Search for a word you can see (Ctrl/Cmd + F). No result? Scanned.
- Zoom in a lot: scanned text gets blurry or pixelated; digital text stays sharp.
How OCR works (in brief)
OCR software cleans the page image (straightening, contrast), finds the text lines, segments characters or words, and recognises them using a model trained for each language. Modern engines like Tesseract use neural networks that recognise whole lines, which improves accuracy, but the input quality still matters most.
Getting accurate OCR
- Resolution: 300 dpi for small print, 200 dpi is fine for normal text. Below 150 dpi, errors climb.
- Flat and straight pages: use a flatbed or a scanner app that flattens and crops; avoid curved book pages.
- Contrast: black on white is ideal; faded thermal receipts and coloured backgrounds are difficult.
- Language: OCR uses language models — documents in English, French, German, Spanish, Italian or Portuguese are recognised best by our tool.
- No handwriting: handwritten notes are rarely recognised reliably.
Summarize a scanned PDF in one step
- Upload the scanned PDFto the scanned PDF summarizer or the home page tool.
- Wait for OCRthe progress shows “Scanned PDF: running OCR”; count a few seconds per page.
- Choose the outputsummary, key points, questions, or the recognised text as .txt.
- Verify critical detailscompare names, amounts and dates with the original image.
Typical OCR errors to look for
- 0 and O, 1 and l and I, 5 and S, 8 and B in numbers and codes.
- Dropped decimal points or thousands separators in amounts.
- Merged columns in multi-column layouts.
- Headers, stamps and signatures read as random characters.
A summary built on OCR text inherits these errors, which is why page references and a quick check of key figures matter.
Frequently asked questions
Can I summarize a photo of a document?
Yes, if it's saved as PDF and the page is flat, sharp and well lit. Scanner apps that crop and flatten give much better results than raw photos.
Why does OCR take longer?
Each page has to be analysed as an image. Count roughly 2-4 seconds per page, versus almost instant reading for digital PDFs.
Is there a page limit for scanned PDFs?
On the free tool, OCR covers the first 40 pages of a scanned PDF. Digital PDFs can have up to 300 pages.
Can I get the OCR text?
Yes, use the Extract text tab or the PDF to text tool and download the .txt.