7.5.5.1   Word Processing

To use word processing, activate Enable on the right side of this dialog. To correct the result of the OCR classification according to certain rules, the assistants' word processing offers two options:

Note that, in contrast to the operators do_ocr_word_*, the entire word is used in the Regular Expression automatically, that is the given expression is surrounded by '^...$'.

Furthermore, the maximum number of characters that will be corrected can be selected via Max. Corrections. For up to this number of characters if the word does not match the expression one or more alternatives, like the second best class and so on are considered, depending on the number set for Symbol Alternatives.

Note that if high numbers of Max. Corrections and Symbol Alternatives are chosen the number of possible corrections may be very high. Therefore, if the number of characters to be read is very high, the number of allowed corrections is internally clipped to 5, 3, or 1 if the alternatives multiplied with the number of characters are equal or greater than 30, 60 or 90, respectively.

The necessary corrections also influence the Score (the more corrections the lower the score) of the result that can be viewed in the Results tab.

For more information on how to use the word processing parameters, please refer to the reference documentation of the operators do_ocr_word_*.