What it means
OCR reads the letters and numbers in a picture of text, such as a scanned invoice, a photographed receipt or a PDF made from images, and outputs text a computer can search and process.
Modern OCR models also keep layout information, such as where each paragraph or table cell sits. Combined with a language model, OCR output can be turned into structured fields like supplier, date and total.
Why it matters for a business
Paper and scanned documents still carry a lot of business data. OCR removes the retyping, and with confidence scores it can flag unclear documents for a person to check.
A business example
Things to watch
-
Accuracy drops with blurry photos and handwriting.
-
Tables and multi-column layouts need testing.
-
Use confidence scores to route unclear pages to people.
-
Handle personal data in documents carefully.