Definition
OCR (Optical Character Recognition)
Also called: optical character recognition, text recognition.
OCR (optical character recognition) converts images of text — like a scanned page or photo — into machine-readable characters.
OCR — optical character recognition — is the technology that converts an image of text, such as a scanned document or a photo of a page, into machine-readable characters you can search, copy and process.
Why it matters for documents
Many PDFs are not “real” text — they are images of pages. Without OCR, the words on them cannot be selected or extracted. OCR makes that text accessible, which is the prerequisite for data extraction from scanned files.
Its limits
OCR returns characters, not layout. It can misread thin characters — a dropped decimal point in a financial figure is a classic failure — and it does not, on its own, rebuild rows and columns. That is why a good document tool pairs OCR with structure detection and flags low-confidence cells.
FAQ
Frequently asked questions
- Is OCR accurate?
- Modern OCR is highly accurate on clean scans but can still misread small or low-contrast characters. For data work, the safeguard is flagging low-confidence cells for review rather than trusting every character.
- Do born-digital PDFs need OCR?
- No. PDFs created from software already contain real text, so their characters can be read exactly. OCR is only needed for scanned or photographed pages.
Related terms
Data Extraction
Data extraction is the process of pulling structured data — rows, fields, tables — out of documents, files or systems so it can be used elsewhere.
Read definition → DefinitionData Pipeline
A data pipeline is an automated sequence that moves data from a source through cleaning and transformation to a destination, usually on a schedule.
Read definition →