OCR is a major mode to assist in the onerous task of proofreading text generated by OCR programs. It’s basically a single-user, localized, Emacs-based re-implementation of the Project Gutenberg Distributed Proofreaders web interface. It looks like this in operation:

The code, and more information, can be found here.
See the ChangeLog for full update info
PGDP auto-edits coming soon.