Understanding How the Full Text Utility Works (Indexer and Services)

The Indexer component of the Full Text utility uses a third party Recognition Engine to perform the actual process of optical character recognition on files in stored in the DocuPhase Repository: extracting any found text. The extracted words or phrases are stored within the FTSData table of Microsoft SQL Server.

You can select a different Recognition Engine in order to change the accuracy and/or speed of the process. Tell me more...

Engine	Description	Available Accuracy vs Speed Settings
		Fast	Accurate	Balanced
MTX	Fast, selectable OCR engine. Supports up to 12 languages Allows a maximum of 64 zones on a single image Offers Omnifont, Draftdot9, and Draftdot24 filling methods Provides Checking Subsystem-based correction. Includes 2 page-level accuracy and speed trade off settings	X	X	X
MOR	Supports 119 languages Allows a maximum of 500 zones on a single image Supports character training to achieve improved accuracy Offers Omnifont, Draftdot24, and OCR-A filling methods Provides Checking Subsystem-based correction Includes 3 page-level accuracy and speed trade off settings	X	X	X
FRX	Optimized for speed Supports 54 languages Allows a maximum of 2,500 zones on one image Offers Omnifont filling methods Provides character training to achieve improved accuracy	n/a	n/a	n/a
Voting 2-Way (PLUS 2-way Voting Engine)	Provides accurate and fast machine print OCR capabilities.	X	X	X
Voting 3-Way (OmniPage 3-way Voting Engine)	Provides accurate and fast machine print OCR capabilities.	X	X	X

---

In general, Windows Services are used to control long-running executable Applications (e.g., Full Text Indexer) that run in their own Windows sessions. In order to process Documents 24x7 and maintain stability, the Full Text Service is used to start and stop the Full Text ndexer after the processing each page of a image/file. One of the main benefits of the Services is that it runs behind the scenes without interfering with other Users who may be working on the same server: continually monitoring the Full Text Queue for Documents that need to be Indexed.