Friday, July 15, 2011

Scanners Live In Vain

For the latest chapter in my scanning saga, I purchased Nuance Omnipage 18, a highly recommended OCR product.

So far this has worked about as well as everything else I've done to try to scan in documents. That is to say, not at all.

Here my problems are simpler, and seemingly insoluble.

First problem: I can't install the software. It goes through the involved installation procedure, tells me it has installed successfully, then nothing. No icon on the desktop, no entry on the list of installed programs, zero.

There's more to the story, of course, including a blown install, a messed up email address (mine, mea culpa) and much fooling with the stuff. All to no avail.

Okay, time to call tech support == and that leads me to the next problem. After playing ring-around-the-rosy on the web site several times, I still can't find the tech support number. Everything seems to indicate there is one, but I can find no trace of it on Nuance's pretty but seemingly useless web site.

As I say Nuance comes highly recommended, but if you can't install the program it's useless.

Now all I've got to do is figure out how to send the damn thing back for a refund. And I'm still without OCR software.

--RC

6 comments:

cracki said...

Adobe Acrobat (not Reader) has a simple OCR function. Works well for me, but there's no interactivity like in OmniPage or FineReader. The Acrobat's HTML export of an OCRed document spits out crazy markup, but it *can* be digested.

If you'd post the raw scans, I'm sure an OCRed version would magically appear in the comments ;)

Rick Cook said...

Thanks Cracki
Right now I'm preparing for the future. I've got Obsidian Harvest OCRed and I'm cleaning it up. (Big job. Way to big.)

Now I'm looking toward the future. I plan on publishing a collection of my short stories later this year and that's going to mean a lot of scanning. I want something better than what I have so far.

George Ellis said...

On a more basic level, and because I have one, I tried our HP OfficeJet 8500A to scan to text and then did a scan to jpeg for a comparison. It was reasonable considering I grabbed a GemCAD for Windows printout. The bottom margin munged and it translated that. You can see why in the jpeg.

www.faceting.biz/faceting_information_links/sharedfiles/Scan_Doc0001.rtf

and

www.faceting.biz/faceting_information_links/sharedfiles/Scan_Pic0001.jpg

The scan also tried to OCR the index numbers in the angle views so did sort of messup the diagrams. But good results... for me.

George Ellis said...

Oh, and that is the stock HP software for the OfficeJet. I did not try a multi-page feed though.

Anonymous said...

I haven't been following the blog in a while, but thought i would add a comment here.

Like @cracki I use Acrobat Standard for scanning business documents. It OCRs them as it goes and makes the text in the document selectable for C-n-P.

I hate to suggest this, but there are several writers books available on Torrents, you might see if someone already did all of the work getting it into text, for you. Then because my experience of reading out-of-print books that way, you will only have to make corrections to the document. It would have likely been scanned by someone with a book scanner and they would have already built a work flow that leads to volume, and at least their piracy would save you some money, even as it looses you sales.

Anonymous said...

ps, to my last post, I'm glad you are feeling better. I personally like paper books over electronic because when the power goes out and we are shoveling cold, a good book under the blankets is a nice way to warm back up if you don't have other options available.