Classification profiles - ABBYY Documentation

ABBYY FlexiCapture supports three classification profiles. Choose one based on whether your documents are visually distinct, differ mainly in text, or need both.

Profile	Use for
Image	Documents with clear visual differences that a person could identify without reading the text, such as passports and driver’s licenses.
Text	Documents that are not visually distinct and must be read to be identified, such as different types of contracts. Uses OCR with statistical and semantic analysis.
Combined	A combination of the Text and Image profiles. Runs in one of two modes, fast or thorough.

Combined profile modes

Fast – FlexiCapture applies the Image profile first, and then the Text profile only for documents that were classified with low confidence.
Thorough – FlexiCapture applies the Image and Text profiles simultaneously.