Another problem with Abbyy OCR

[rihanna27]

Hay,
now i have i a much bigger problem.I dont know how to set that scaned documet with output format html,xml,pdf can give me only data which i want,like bill,supplier,date of bill and so on. I dont know how to define types for documet,because problem is that those datas are always on diffrent place.I think this is problem with position of datas and i dont know how to solve it.


Any ideas?

[devils casper]

There must be instructions for setting output type of scanned docs in Abbyy documentation. Have you checked it thoroughly?

[rihanna27]

I think we didnt understand each other, i know how to set output types,but i dont know how to select on scan document only metada which i need.I found that is related with setting block types but the rest i is still a dark.

[gogalthorp]

Let use try to understand what you want.

You want to scan invoices from multiple sources which are formated differently. You wish to only pick out certain data (date amounts etc) which are located in different places on the different documents.

Do you know of any software that will do this on any OS??

It seems to me that the software first would need to recognize a given format and you would need to associate that format with the locations of the wanted datum. Each new invoice source would need to be taught to the system. That also assumes that each format could be recognized in the first place by some unique feature.