Recognition module
The Engine can load several recognition engines. Licensing determines which ones are available. You can select which available engine to run in a given user zone by specifying a particular filling method suitable for the zone content, or by specifying an actual recognition engine. In addition, you can also set which recognition module will be used by the Engine for the processing of the individual OCR zones that are created automatically. If no instruction or guidance is given, the recognition engine is selected automatically.
The following recognition engines are available:
-
PLUS3W: 3-way voting omnifont engine for machine-printed text
-
PLUS2W: 2-way voting omnifont engine for machine-printed text (default)
-
MTX: M/TEXT omnifont engine for machine-printed text in 12 languages
-
MOR: MOR multi-language omnifont engine for machine-printed text in over 120 languages
-
FRX: Fireworx multi-language omnifont engine for machine-printed text in 56 languages
-
DOT: engine for 9-pin or 24-pin draft dot-matrix printouts
-
MAT: matrix matching engine for codified scripts (for example, OCR-A, and OCR-B)
-
HNR: engine for hand-printed digits
-
RER: re Recognition (third-party) engine for hand-printed alphanumerical characters
-
OMR: engine for optical marks, for example checkmarks
-
BAR: engine for bar codes
PDF recognition
PDF files can be separated into two basic classes, image-only PDF and normal PDF. Recognition of image-only PDF is exactly the same as that of any other image format. The normal PDF contains both image and text data. The CSDK can extract this textual information to boost OCR accuracy. In general, text in a PDF file is reliable, so accurate results are possible even when the quality of the imaged text is low. See New features and changes for improvements in PDF handling. The fullest description of PDF functionality in the CSDK is found in the Related Pages section of the RecAPI online help system.