We have some clients using ImageRight - http://www2.imageright.com/. It is focused on insurance.

And I have written a document management system myself for a large insurance company that has 4 scanning workstations that take all the correspondance received in their mail room and store it on disk. It is VFP with a SQL backend. It uses ezTwain for the scanning and uses the image viewer from Viscom - http://www.viscomsoft.com/imageviewer.htm.

The scanners were Ricoh IS300's.. http://shopping.yahoo.com/p:Ricoh%20IS%20300e%20Scanner:1991731126



>I have a client interested in reducing paper at their office. Going completely paperless is probably unrealistic but they'd like to see what's possible, both to reduce costs and be good for the environment.
>
>They have about 20 people on a LAN, they're not especially heavy paper generators, I'd say about average on a per-head basis. They'd like to start scanning paper documents, so we're looking for recommendations for a suitable scanner and software.
>
>Sharing a scanner attached to one user's workstation is not acceptable, due to inconvenience to that user. At this point we think we're looking at a networked "walk-up" solution of some kind (same idea as a photocopier). The way we'd like it to work:
>
>- Load in the document to be scanned. The scanner should have an Automatic Document Feeder (ADF) able to handle (say) 20 pages without requiring hand feeding (i.e. "feed & forget").
>- Being able to handle legal paper would be nice
>- Having flatbed platen capability (like a photocopier) would be nice as well, for photocopying books, manuals etc. However, I suppose those could be photocopied, then the photocopies scanned. So, a sheet-feed-only scanner is not out of the question
>
>- Select output type. In addition to images like .JPG or .PNG, the scanner should be able to output to .PDF and Word .DOC.
>
>- Select destination. This would likely be a folder share on a server somewhere. It looks like some systems let you send output to e-mail addresses, which is an interesting option but we probably want to avoid clogging up Exchange with all this stuff.
>
>In my so-far limited research I've seen some networked scanners that can do this. However, my biggest question is how well the scanner can create .PDFs and .DOCs. In both cases, OCR or other processing software can potentially "cheat":
>
>- some cheap PDF solutions just take the raw images and create a .PDF out of a series of these images. Since they're images, you can't search them for specific text etc. I get the impression these files are larger than "true" .PDFs and since their included images are a fixed resolution, they don't generally look as good - can't take advantage of scaleable fonts, etc.
>
>- I've almost always been disappointed by convert-to-Word features of various OCR packages. For the simplest documents (little or no graphics or formatting) they can work OK but their (spelling) error rate still seems high. Documents with significant formatting or graphics tend to be either completely munged, or else the OCR cheats by using a lot of text or graphics objects in absolute positions. In this case the resulting document can't be edited or updated, it has to be recreated. The goal (holy grail?) is a properly formatted Word document, mostly text and with graphical elements only where necessary (where they would have been required on a fresh document, anyways).
>
>I'm convinced that OCR performance for .PDFs and .DOCs is absolutely critical to the success of scanning in a reduced-paper office. If it works well, it'll be used a lot. If it doesn't, it'll just sit there gathering dust.
>
>It may be that the only way to get this kind of software/OCR reliability is to go with a full-blown document management system. If so, that's not out of the question, but cost is always an object :)
>
>So, if anyone has any experiences or recommendations for scanning systems with really good OCR/conversion capabilities, I'd love to hear about it.