Proofreading

After a page is recognized, the recognition results appear in the Text Editor. Proofreading starts automatically if that was requested in the Proofing panel of the Options dialog box. You can start proofing manually any time the program is not busy. The Proofreader will stop on suspected words marked with colors. Marked words are the ones that will be offered for inspection during proofing. Color markers are removed from words in the Text Editor as they are proofread.

Markers

Marking is done with red wavy underlines and dark yellow highlights. Red wavy underlines indicate both non-dictionary words and suspect words. Dark yellow highlights indicate suspect characters.

Non-dictionary words: These are words that were recognized confidently, but are not found in any active dictionary: standard, user or professional. These appear only in languages with dictionary support. To see which languages have built-in dictionaries, click here.

This is an example for a non-dictionary word:

Omnipage eng non dict%20word Proofreading

Non-dictionary words will be marked if the Mark all non-dictionary words option on the Proofing panel was enabled. Words are likely to be marked, even if the recognition engine produced the result with high confidence. This marking will also appear on text entered or pasted into the Text Editor after recognition. If this option is turned off, proofing will stop only on suspect words, marked with red underlines and possibly dark yellow highlights.

Words with suspect characters: These are words with unrecognized characters or dictionary-approved words containing one or more characters recognized with lower confidence.

This is an example for a suspect character:

Omnipage eng suspect%20char Proofreading

Unrecognizable characters are represented by a red reject character. (A tilde ~ is the default reject character — define a different character in the OCR panel of the Options dialog box.)

Suspect words: These are likely to be non-dictionary words with one or more suspect characters, but may also be suspect for other reasons. Non-dictionary words recognized with high confidence by all recognition engines or appearing repeatedly in a document may not be marked red.

This is an example for a suspect word:

Omnipage eng suspect%20word Proofreading

You can choose to show or hide the markers in the Text Editor panel of the Options dialog box. Whenever you start proofing or choose Find Next Suspect the markers are turned back on. The markers are not exported with the text, but they are saved to the OmniPage Document (even when hidden).

  • It is a good idea to put your name, address elements, and your most common trade, company and product names or abbreviations into a user dictionary, so they do not slow down proofing.

To proofread OCR results and correct errors:

  1. Click the Proofread OCR button Omnipage tb st proofread Proofreading or press F4 to find the next suspect word without displaying the Proofing panel.

  2.  

    • If markers were hidden in the Text Editor when proofing is started or Find Next Suspect is chosen, the markers become shown and remain shown after proofing.

     

    • A page is marked with the proofed icons Omnipage icon proofed Proofreading and Omnipage icon mod proofed Proofreading on its thumbnail and in the Document Manager if proofing ran to the end of the page.

     

    • Proofing should not be used for Japanese, Chinese or Korean. Disable automatic proofing under Options / Proofing.

     

     

Proofreading