What is OCR? How to extract text from a scanned PDF
A scanned PDF is just a photograph of a page. The letters you see are pixels — the PDF does not know they represent words. OCR changes that.
What does OCR stand for?
OCR stands for Optical Character Recognition. It is the technology that analyses an image and identifies characters, turning a picture of text into actual, selectable, searchable text.
OCR is what happens when you photograph a receipt and your banking app reads the amount automatically. It is also what allows Google to search inside scanned books, and what courts use to digitise decades of paper records.
When do you need OCR?
- You scanned a contract or invoice and cannot select or copy any text
- You received a PDF that opens as a series of images rather than text
- You want to search inside a document but the search returns nothing
- You need to extract data from a scanned form or table
How OCR works
Modern OCR engines (including the open-source Tesseract engine used by AmorePDF) process an image in several steps:
- Pre-processing — the image is straightened, de-skewed, and its contrast is enhanced
- Segmentation — the engine identifies regions of text, separating them from images and white space
- Character recognition — each character is compared against trained models to identify the most likely match
- Post-processing — the output is checked against language dictionaries to correct errors
The result is a text layer that sits over or replaces the original image inside the PDF.
How to extract text from a scanned PDF
Our OCR PDF tool runs the entire recognition process in your browser using WebAssembly. No file is ever uploaded to a server.
Step 1 — Open the tool
Go to OCR PDF.
Step 2 — Load your scanned PDF
Drop the file onto the upload area. The tool loads a preview of the pages.
Step 3 — Select the language
Choose the language of the document. Accuracy improves significantly when the correct language model is selected. Common options include English, French, German, Italian, Spanish, Portuguese, and many more.
Step 4 — Run OCR and download
Click Extract text. Processing takes a few seconds per page. Download the result as a plain text file (.txt).
OCR accuracy: what to expect
Accuracy depends on scan quality:
| Scan quality | Expected accuracy | |---|---| | Clean, high-contrast (300 dpi+) | 98–99% | | Good quality (200 dpi) | 93–97% | | Low contrast or skewed | 80–92% | | Handwriting | 50–85% (varies widely) |
Printed text in good condition is recognised with very high accuracy. Handwriting is harder and results vary by style.
OCR vs. extracting existing text
If your PDF already contains a text layer — it was created digitally, not scanned — you do not need OCR. Use Extract Text instead, which copies the existing text directly without any image processing. It is faster and produces a perfectly accurate result.
Not sure which one you need? Open the PDF in a browser and try to select some text. If you can select and copy it, use Extract Text. If the cursor does not grab any text, the document is a scan and needs OCR.
Other OCR tools
- OCR Images — run OCR on image files (PNG, JPG) rather than PDFs
- Extract Scanned Text — optimised for multi-page scanned documents
Conclusion
OCR is a powerful but often overlooked tool. It turns a static image into a living document you can search, copy, and edit. AmorePDF runs OCR entirely in your browser — privately, for free, with no limits.