Configuring recognition and correction, Overview, Recognition configuration methodology – Kofax INDICIUS 6.0 User Manual
Page 39: Chapter 5
Getting Started Guide (Fixed-Form)
33
Chapter 5
Configuring Recognition and
Correction
Overview
The following tutorial shows how the Definer tool can be used to generate a
definition file (.idf) for Recognition and Correction. Definer allows configuration
changes to be made quickly. It includes an embedded Recognition engine that allows
the effect of changes on the recognition process to be tested immediately.
The definition file created here specifies a subset of the data fields on the example
images used in the
.
Note
This definition file will use only a subset of available features of Recognition
and Correction. This subset is generally common to all definition files. For details of
how to configure all features of Recognition and Correction, see the INDICIUS Help.
Recognition Configuration Methodology
Recognition can capture many different kinds of data from document images to a
high degree of accuracy. It can also process batches of mixed document types,
automatically determining the type during “registration”.
The typical process to recognize a document (which Recognition does every time it
runs) is as follows:
Step 1: Identify the document type.
Step 2: Correct for misalignment.
Step 3: Locate the data fields.
Step 4: Remove noise from the data fields.