what am I doing wrong?

  • Last Post 2 weeks ago
samehnassar39 posted this 3 weeks ago


From Java program, I use ABBYY Cloud OCR to recognize tiff invoices, I use receipt mode which will return xml file with the invoice information. In xml I can see <recognizedItems> which will return invoice line items. I have some issues in this:

- I can't find all invoice items picked in xml (e.g invoice has 10 items it pick only 2 or 3) and I can't find item code which is the most important part.

- When I send tiff file with multiple pages, OCR will pick the items in the first page only and ignore rest of pages (is there any configurations I should do to make OCR pick all data in all tiff pages).

Is there something wrong I do?

Is receipt mode is the best option to recognize invoices? if not what is the best options I can use from ABBYY Cloud OCR to recognize invoices?



Oksana Serdyuk posted this 2 weeks ago


Sorry for the delay, we have answered you by email in details. Briefly, the processReceipt method is appropriate only for receipts processing and only for one-page files. The most reasonable will be to use the full-page recognition mode (the processImage method) and then apply some post-processing for extracting necessary data.