Problem Description:
When attempting to "Recognize Text", one of the following error messages appears:
Problem Origin:
It seems to be a compatibility issue in the PDF and not in the OCR algorithm itself, although this is speculation and not an official statement.
Problem Workaround:
A possible solution consists of saving the afflicted PDF pages as PNG images (lossless), opening the PNGs in Acrobat, and then performing OCR. Specifically,
This has worked for every problem instance that I've come across, and there appears to be no quality loss. If anyone else can verify this solution, I would appreciate it.