OCR (Optical Character Recognition)
Technology that extracts machine-readable text from scanned images or image-based PDFs, making content searchable and selectable.
Last updated
What Is OCR?
Optical Character Recognition (OCR) is the process of converting an image containing text — such as a scanned document, photographed page, or image-only PDF — into machine-readable, searchable, selectable text.
Why Many PDFs Need OCR
When a physical document is scanned, the scanner creates a photograph of the page. The PDF contains only pixels — no actual text data. You cannot:
- Select or copy text
- Search the document with Ctrl+F
- Have a screen reader read it aloud
- Reuse the text in another document
OCR analyses the pixel patterns in the image, recognises character shapes, and embeds the recognised text as a transparent text layer over the image.
OCR Accuracy Factors
| Factor | Impact on Accuracy |
|---|---|
| Scan resolution | 300 DPI minimum; 400 DPI recommended for small fonts |
| Page skew | Tilted pages reduce accuracy significantly |
| Font clarity | Clean printed text: 99%+ accuracy; handwriting: 80–90% |
| Language | Most OCR engines support 100+ languages |
| Image quality | Low contrast, noise, and shadows reduce accuracy |
When OCR Is Needed
✅ Scanned contracts, tax forms, or invoices that need text extraction ✅ PDFs where Ctrl+F finds nothing ✅ Documents received from government agencies or older offices ✅ Photographed receipts or pages from books
When OCR Is Not Needed
Your PDF already has selectable text if you can click and drag to highlight text in any PDF viewer. If highlighting works, OCR has either already been applied or the PDF was created digitally (not scanned).