How It Works
OCR (optical character recognition)
Software that turns images of text — a scanned page, a photo of a sign, a PDF made from a scan — into text a computer can search, copy and edit. Modern OCR uses machine learning, and vision-language models can now read text in images as part of answering questions about them.
Origin · no single documented coiner
Machines that read printed characters date to the early 20th century (Emanuel Goldberg’s and Gustav Tauschek’s patents in the 1910s–1930s); Ray Kurzweil’s 1970s reading machine for blind users popularized OCR that could read any typeface. No single coiner of the term.