Optical Character Recognition (OCR) meaning

In legal practice, optical character recognition (ocr) is the software process that converts scanned or image-based documents (for example, PDFs or TIFFs) into machine-readable text so they can be searched, indexed, reviewed and redacted for disclosure, discovery and investigations. In England and Wales, the term was defined in former cpr Practice Direction 31B para 5(9) as the computer‑facilitated recognition of printed or written text characters in an electronic image in which the text cannot be searched electronically. Although PD 31B has been revoked, the description remains in common e‑disclosure/e‑discovery use, including under PD 57AD (Disclosure in the Business and Property Courts). In Scotland, Northern Ireland and Ireland, OCR is a descriptive technical term used consistently in productions/discovery practice and is not generally defined in rules or statute. Key features and practice points: - Creates a searchable text layer for image‑only documents; it is not a substitute for native electronic text. - Accuracy depends on image quality and layout; handwriting (ICR) is less reliable and may require quality assurance. - Typically used to enable keyword searching, analytics, deduplication and technology‑assisted review. - Parties commonly address OCR scope, settings, quality thresholds and exception handling in disclosure/discovery protocols, while preserving originals.