2.1. Stage 1: Document type detection
Configuring document type detection allow the application to apply existing extraction/conversion configuration to next documents of the same type.
For document type dectection we need:
-
A name for the document type.
-
Some text selection in the document that make it unique:
-
Text to idenfity the document generator:
-
Supplier vat number, supplier name or account number
-
Text to identify the document layout:
-
Text from grid headers
-
Text from forms
Usually, two or three line detection elements is enought to automatically detect the document type.