Greetings,
> Davis, Tom wrote:
>
> I agree that for the vast majority of scanning applications, scanning until
> the ADF is empty is exactly what the user wants. However, there are common
> variations...
>
> In the production scanning world, group of pages to be scanned are sometimes
> interspersed with separator pages on which information about what to do with
> the group of pages is encoded. The scanners can actually decode the separators
> and process the group accordingly.
And sometimes the document itself has a barcode on it that holds a key piece of
information which lets you know how to "index" the information into a database.
In each case, there needs to be a way to get this information from the decoder
(in the backend) to the indexer (in the frontend). The way that I do that with
scanadf and the Bell+Howell is with a TEXT frame type. Instead of getting image
data, you get an xml encoded text that has all the info the the decoder wishes
to send. Then the indexer parses it and does what it might, such as changing
pagination and/or document numbering or assigning a keyword to the image(s) for
storage in a database.
Tom
-- Source code, list archive, and docs: http://www.mostang.com/sane/ To unsubscribe: echo unsubscribe sane-devel | mail majordomo@mostang.com
This archive was generated by hypermail 2b29 : Thu Mar 01 2001 - 17:35:07 PST