• The VOIDRUNNER'S CODEX is LIVE! Explore new worlds, fight oppressive empires, fend off fearsome aliens, and wield deadly psionics with this comprehensive boxed set expansion for 5E and A5E!

Anyone know anything about Imaging documents?

reapersaurus

Explorer
My work scans their Repair Orders, then manually records the work order # for tracking purposes in the Comments field of the file.
This comments field is visible to the 98 clients, but invisible to the XP clients.

The scanner isn't working very well at all anymore, and they've stopped using it.

I need to find a substitute hardware and maybe improved procedure for this tracking system.

Anyone have any recommendations for a relaible, fast scanner and maybe an OK OCR system?

Thanks.
 

log in or register to remove this ad

Rackhir

Explorer
What kind of scanner and OCR software have you been using? Also what kind of throughput do you need per day? Finally, what kind of system are you using to process the ocr with? That can be quite processor intensive.
 

reapersaurus

Explorer
We've been using a Fujitsu ScanPartner 15C (using Kodak Imaging Professional software) :eek:
We're not currently using OCR.

The Fujitsu is not being used anymore because it was skipping pages too often.

I'm guessing they get at least a couple hundred pages a day they need scanned.

The OCR is something we'd like to do in the future, if it would be easier to use than the scanning person just typing in the #'s in the Comments field.... :D

We're a small company, in the big scheme of things, and aren't high-tech at all (I'm new here) ;)

Any help would be appreciated - we'd prefer not to have to send the scanning off-site, to a 3rd party.
 

Sir Whiskers

First Post
I may be missing something, but if you're currently just scanning the pages as images, take a look at the copier your company uses (assuming you have something larger than a tabletop model). If you're leasing from a vendor, there are a lot of decent digital models on the market that will handle your copying, work as a digital printer, fax documents, and scan pages.

A few things when talking with the vendor(s):

1. What is necessary to connect the device to your network? Things are much better now than before, but I still remember having fits with one brand of copiers (actually the fiery software they used) because of security settings used on our network wouldn't allow the device to connect.

2. In what format will it scan the images? You never, never, never want to use a copier/scanner that uses a proprietary format. That just locks you into that company's products and if they ever stop supporting the format, you're almost certainly SOL.

3. Where will the images be saved? How easily can you access them from your desktop after scanning? If special software is needed to view the images, only pursue it if the software if freely and perpetually available (IOW, if you ever stop leasing from the vendor, you can still access your old images).

4. Always test the exact model you're considering, preferably on-site. Most good vendors will bring out a copier for a test, some for a few days, some for longer. This allows you to test it in your own real-world environment, as opposed to the vendor's. If on-site isn't an option, at the very least test several types of documents at the vendor's showroom.

As far as I know (it's been a while since I handled copiers for a company) you won't have OCR capability with these devices - they scan as an image which (IIRC) cannot be modified. But since the features on these machines are constantly changing, it's probably worth your time to at least speak with a couple vendors about it. Worst case, you just waste a bit of time finding out they don't have what you need.
 


reapersaurus

Explorer
Thanks for the recommendations.
But I think my company would prefer to stick within the technologies that they have experience with (scanners), instead of utilizing one they don't know (copiers, senders).

Does anyone have experience with modern scanners and OCR?
 

Breakdaddy

First Post
I have OCRd several things here at work on an HP Scanjet with ADF. It works nicely and quickly. My OCR software of choice is Omnipage Pro 14, which (when using 300-600 DPI on scans) yields 98%+ accuracy on most documents. Hope this helps. Good luck!
 

Rackhir

Explorer
I've done some looking and that Fujitsu you've got is supposedly one of the best of the affordable scanners with an ADFs. Perhaps you just need to replace the ADF or get it repaired? Most of the other models are multiple thousands of dollars, would your company be willing to spend $2,000-$5,000 for a new one?

If all you really need on the documents is the work order #, OCR is unlikely to do you much good. Do they already have the work order # on them and if so is it typed or handwritten. If it's hand written, it's unlikly that OCR will work terribly well unless your people have very neat and consistent handwriting.

I haven't had much experience with this, but I would suspect that the OCR would add at least a minute or so to the processing of each document and if you are doing several hundred per day, that can really add up. Perhaps Breakdaddy can give us an approximate time per scan for the OCR to operate? Please tell us what kind of machine you have, since that has a strong impact on how fast the OCR software would run.

Finally with regards to the comment field's visibility, check the version of "Acrobat Reader" that the 98 and XP clients are using. I believe that the scanner saves the files in PDF format and Reader is what they would be using to open those files.
 
Last edited:

frugal

First Post
Rackhir said:
I haven't had much experience with this, but I would suspect that the OCR would add at least a minute or so to the processing of each document and if you are doing several hundred per day, that can really add up.


OCR packages like Omnipage PRO let you work in batch mode with what they describe as templates.

You use templates to define where on the page you are interested in text. So if you were scanning in work orders you would set up a block for each piece of text you wanted scanned in: Order number, address, etc and give each block an ID

Then the scanning software just scans those blocks and has already been told what sort of data is in each of the blocks so it takes a very short time to OCR. The OCR software will then spit out a text file, XML file CSV file, whatever, with the scanned text in it so that you can automatically pass the data to your database (or whatever the next step of the process is).
 

Breakdaddy

First Post
Once scanned I can get a page fully OCRd in less than 15 seconds (if the scan was high quality 300+ dpi). I do this on my work machine which is a P4 3.02 ghz
800mhz fsb hyperthreading Dell. Its also got 2 gigs of ram and dual SATA HDDs so its far from typical. At home I have an athlon 3200+ with 512 ram and it only takes a few seconds longer, maybe 20-25.
 

Voidrunner's Codex

Remove ads

Top