Optical Character Recognition (OCR) usually is akin to the joke about the talking dog, "It's not that he does it so well, but that he does it at all."
And then along comes OmniPage Pro 11.0 from ScanSoft. Last year, I looked at the Peabody MA-based OCR giant's mid-market Pagis Pro Millennium Scanning Suite last year (TNPC #3.11). I wanted to check out its professional OmniPage product lineup, added when ScanSoft acquired rival Caere, Inc. A company spokesman suggested I hold off until Version 11 hit the market--and now I see why.
The bottom line in determining the value of an OCR program is very simple: by the time you get through fixing recognition errors, did the process take less time than re-typing yourself would have? And did it make fewer typographical errors?
Very simply, the answer for OmniPage is "absolutely yes"-- bordering on "oh wow!" It is the most accurate recognition engine I have ever used. On good to moderate quality documents, it makes almost no text recognition errors. I rarely saw more than two on three per 250 words, and most of those were caught by its internal correction system.
Note I said, TEXT recognition. OmniPage's "TruePage" technology, which attempts to make the layout of the electronic document match up with the hard copy, does a fine job. But think of the output as a good first draft. You are going to have to manipulate the electronic file to make things exactly right. Also, I found that some tables did fool the program, especially when borders were done in something other than solid rules.
The killer addition to OmniPage 11 is its ability to scan Adobe Acrobat PDF files and turn them to editable Word, html, text, etc. files. It allows you to use--and hey, let's not be violating anyone's copyrights, out there--excerpts from reports that were published in PDF, convert to a format with a smaller file size, and generally make the material easier to work with. OmniPage, of course, also has all the expected bells and whistles: Office integration, the ability to clean up faxes, and so on.
For whatever marketing reason, ScanSoft has a theoretical $500 list price
for the full version of OmniPage 11, But you have no need to pay
anything close to that. The upgrade version has an affordable street
price of $130:
http://www.TheNakedPC.com/t/415/tr.cgi?alocr2
As ScanSoft spokeswoman Jane Van Saun points out, "anyone with ANY previous OCR software can upgrade ... And almost anyone with a scanner has some type of OCR bundled with it." Also, the upgrade is available to users of Microsoft Office XP, which uses the OmniPage OCR engine in its OCR utility.
If somehow you still aren't eligible, PaperPort Deluxe 7.0 has been
released and for $55 you get the state of the art in document management
software plus the upgrade rights to OmniPage.
http://www.TheNakedPC.com/t/415/tr.cgi?alocr3
You can reach Al Gordon at:
mailto:al@TheNakedPC.com
