I'm not sure if this is even possible, especially since there is
barcode recognition involved. Here's what I'm looking for:
I want a program to be able to point to a directory of single page
*.TIF files, some of which contain barcodes. Each barcoded page
determines the file name for the files following that page named for
the barcode itself. In other words....
Page 1 - Barcode of "12345"
Page 2 - random page
Page 3 - random page
Page 4 - random page
Page 5 - Barcode of "67890"
Page 6 - random page
Page 7 - random page
Page 8 - random page
Running the program in question would result in 2 multi-page TIF
files: the first named 12345.tif containing the images for pages 2-4;
the second named 67890.tif containing the images for pages 6-8. The
barcoded header sheets would be discarded. Make sense? You can see a
sample of what the barcoded sheets look like here:
http://www.baileydakota.com/sample_barcode.tif
Using the sample_barcode.tif file example, the pages following it
would be contained in a multi-page tif file called
WO264170-0482561447.tif, the first page being the first page following
the barcoded header sheet. I am looking to do this in large
batches...a directory may contain 2,000 individual sheets including
the barcoded header sheets. The program would need to ask for a
source directory (where the raw, single-page files would be), and a
target directory (where the final, assembled multi-page files would
go).
Is it possible? Can it be done for $50? Am I totally insane?
Thanks for looking. |
Request for Question Clarification by
leapinglizard-ga
on
07 Sep 2004 09:40 PDT
It's possible, yes, but difficult. Your task entails some pretty
heavy-duty image processing and optical recognition. We're talking
artificial intelligence and advanced math. Researchers have most
likely dealt with this problem before, but I doubt you'll find a
software package ready-made to tackle your particular situation. If
you wanted custom software made, you would have to hire a computer
scientist specializing in image processing and at least one developer.
They would take one to three months to complete the project. I'd put
the R&D budget somewhere between $50,000 and $250,000. Unless you have
millions of files, you'd be better off hiring a clerk to do the job
manually. Sorry to be the bearer of bad news.
leapinglizard
|
