OCR (Optical Character Recognition)
OCR (Optical Character Recognition) converts the text present on an image or a scanned document into characters a program can use. It comes into play as soon as a supporting document arrives as a photo or an unstructured PDF: identity document, bank statement, invoice, tax assessment notice. Two limits to be aware of. The error rate depends heavily on the quality of the capture, and a single character misread in an IBAN or a document number is enough to distort a check. Above all, OCR reads what is written without saying whether the document is authentic: extracting the data from a forged document works perfectly well.