The Google Docs Blog announced you can now upload old scanned documents into Google Docs and Google will OCR the documents. OCR will convert the text scans from a picture into selectable and searchable text within Google Docs, which you can then export to a multitude of file types.
Back in October, Google would OCR web documents that were scans and make those searchable. But there was no easy way for a normal user to scan in their own document and have Google OCR it outside of using the Google API. Now anyone can go to Google Docs, click upload and have Google convert the scans to text for you.
Currently, the supported language for this method of OCR are English, French, Italian, German and Spanish but I am sure Google will continue to add more over time.