>>>Hi,
>>>
>>>Our catalog department has a ton of PDF files ... one file for each page of the catalog.
>>>
>>>We need to build a database of which items are on which page and the fastest ( and most accurate ) way to do this would be to read the PDF file directly...
>>>
>>>I've seen PDFs mentioned here before but I'd like to know if anyone's done this successfully and are there any hidden issues.
>>>
>>>TIA
>>
>>We do OCR through 'Adobe Capture' and have seen great results. Adobe is boasting 98% readability.
>>
>>I work with a guy who automates Adobe for our Document Imaging System.
>
>When you say OCR, do you mean you run the pdf files through the "Adobe Capture" software package?

Talking to the programmer he said you can go to the Adobe web site and download their system development kit for Capture. You then can look through their examples and help files one what it can do.

The basic thing they do here is this:

Someone sends in a PDF or set of PDFs.
The PDF is put through Adobe Capture using automation. The automation has parameters that tell it how deep OCR it will do (full text searching usually).
The Capture puts out a PDF with a searcheable index for the document.
You can use automation to look for certain things on the document through the index that Capture built.

He also pointed out that you can Capture and OCR a group of PDFs at a time.