Just came across this (haven't tried, but seemed worth sharing).

OCRmyPDF combines several Linux tools to generate PDF/A files (PDF format intended for long term archiving) from PDF files that contain just bitmaps with text, e.g., of scanned document pages. The generated PDF/A files are then searchable.

Original article in German describing the tool: http://www.heise.de/open/artikel/Toolbox-Texterkennung-mit-OCRmyPDF-2356670....

Google translation: http://www.google.com/translate?hl=en&ie=UTF8&sl=auto&tl=en&...

Project homepage on github: https://github.com/fritz-hh/OCRmyPDF

Cheers, Peter