Tesseract is popular because it is open source, in my opinion. I've had a great deal of frustration using it, but to be honest I think OCR falls into the same valley as trying to implement an office suite, or reimplement the Windows api - high expectations and only so many spare cycles to work on it.
I also have FineReader for Mac but I've had contact with some of Abbyy's more expensive stuff and it's really incredible. I wouldn't recommend any other system, if asked.
I also have FineReader for Mac but I've had contact with some of Abbyy's more expensive stuff and it's really incredible. I wouldn't recommend any other system, if asked.