Conjecture::StandardProcess Class Reference

#include <StandardProcess.h>

Inheritance diagram for Conjecture::StandardProcess:

Conjecture::ProcessComponent Conjecture::Component Conjecture::Root List of all members.

Detailed Description

Implements the most common 'process' algorithm.


Almost every OCR will use the same 'process' algorithm, and will instead vary in how the cub-algorithms within 'process' are implemented. In particular, the 'process' algorithm almost always consists of

1) Elementitioning the image into Elements (Glyphs, Words, Lines, Regions). Note that only Glyph partitioning in strictly required during this phase.

2) Identifying the unicode character associated with each Glyph. It is not necessary (but may be useful) for the Glyph to already be categorized within a Word, Line, or Region within the Page.

3) Formatting the unicode characters associated with each Glyph. It is at this level that Glyphs MUST be organized into Words and Lines (and possibly Regions) within the Page, in order to ensure that spaces and newlines are inserted in the proper places. However, as mentioned above, identification of Words, Lines and Regions may have occurred earlier as well.

This class implements the standard 'process' algorithm discussed above.

Public Member Functions

virtual bool execute (Page *page)

The documentation for this class was generated from the following files:
Generated on Thu Jun 15 19:56:10 2006 for Conjecture by  doxygen 1.4.6