TALK ABOUT CHARACTERS

We tested the two leading OCR programs against each other and as these things often turn out, each is better at some things than the other.

OCR is tech-talk for Optical Character Recognition and it refers to software that can read a page and keep its formatting. The “read” part is critical because if you just take a picture of a printed page, you can’t edit or search it; the computer doesn’t recognize the words. OCR software understands the words – at least most of them.

The two top programs in this field are Abbyy FineReader, version 9, and OmniPage 16. Abbyy FineReader, Corporate Edition, has a list price of $600,  as opposed to $500 for OmniPage Professional. But Abbyy has a free trial FineReader 9 version, which you can download from their web site: abbyy.com. The program supports 179 languages, reads bar codes and understands headers and footers. There’s also a $400 version of the Abbyy program and a $150 version of OmniPage.

We pause here for a moment to note that we have reviewed such programs for more than 25 years and both of these are by far the best we’ve ever seen. You can’t go wrong.

Tech support is free at Abbyy but $20 per “incident at OmniPage, though the OmniPage 16first call is free. OmniPage supports more than 100 languages and the expensive professional version has  features that will justify its higher price for some users. One feature is form recognition. If the spaces in a business form have been filled out with typed information, the program can read that information and transfer it to a spreadsheet or database. Another is the ability to black-out, highlight or strike through text. We thought the $150 version, however, would suit most users.

While both these programs were created initially to work with scanners, we decided to use them with a plain digital camera – an old two-megapixel Nikon, in fact. We did this because such camera scans are becoming common and are very handy. Going to a library or simply browsing a newspaper or magazine, you often don’t have a scanner nearby. But if you carry a small digital camera, you can simply take a picture of the page and either one of these programs can read that image. You can take pictures of book pages and move those directly into Microsoft Word or most any other text editor. It’s extremely popular with students and researchers.

We took a picture of two pages from the travel writings of Anthony Trollope. (We don’t want to get too obscure here but that’s what Joy happened to be reading.) The print was small but the old camera handled it clearly. With OmniPage we had to say the image was from a camera, and then click “set” to adjust the brightness. If we did not brighten the image, OmniPage made a lot of errors. Once brightened, however, there were no errors. Abbyy’s program automatically detected that our image came from a digital camera and adjusted the brightness. It also read the pages without error.

Both these programs do something that lawyers will like: they maintain the reference numbers in plea briefs. Plea points are numbered, and if the document is scanned and later edited, the program will keep the numbers with the pleas, even if those have changed size. This should also work with script writing, where dialogue is locked to a character’s name. Abbyy’s web site is Abbyy.com and the web site for OmniPage is Nuance.com.

Since some people are bound to wonder about Abbyy’s odd name, we asked and the company says it is a phonetic rendering of a glyph that means “keen eye” or “sharp vision” in an ancient Babylonian language.  (How do they know how a word was pronounced in ancient Babylonia?)

Note: If you are upgrading from a previous version of  FineReader, you can get FineReader9 for $100 instead of $180. Use the coupon code “oncomp” at abbyy.com .

 

Comments are closed.