OCRedited

OCRedited is an open-source, community-driven project that provides a workflow for post-processing OCR outputs to improve accuracy and reliability of digitized texts. It combines automated corrections with human-in-the-loop editing and maintains an audit trail of changes.

The software is designed to work with various OCR engines, including open-source options such as Tesseract and commercial engines, allowing results to be edited and validated within a single interface. It supports import of OCR results, batch editing, and export to editable formats.

Key features include a side-by-side editor that flags suspected errors, dictionaries and language models, spell-checking, version control for edits, role-based review queues, and exports to TEI, XML, JSON, or plain text.

Use cases: digitization projects in libraries, archives, and research institutions, where accuracy is crucial for searchability and accessibility.

Development and licensing: OCRedited is developed openly, with contributions via public repositories; licensing is commonly permissive to encourage adoption; governance is community-driven with transparency in changes.

See also: optical character recognition and post-editing workflows.