Nuance scansoft omnipage 15 Manual De Usuario

Try these solutions if any part of the original document is not converted 
to text properly during OCR:

◆ Look at the original page image and ensure that all text areas are 

enclosed by text zones. If an area is not enclosed by a zone, it is 
generally ignored during OCR. See the section on creating and 
modifying zones, “Working with zones” on page 40.

◆ Make sure text zones are identified correctly. Reidentify zone 

types and contents, if necessary, and perform OCR on the 
document again. See “Zone types and properties” on page 39.

◆ Be sure you do not have an unsuitable template loaded by 

mistake. If zone borders cut through text, recognition is 
impaired.

◆ Adjust the brightness and contrast sliders in the Scanner panel 

of the Options dialog box. You may need to experiment with 
different settings combinations to get the desired results.

◆ Use the Image Enhancement Tools to optimize your image for 

OCR.

◆ Check the resolution of the original image. Hover the cursor 

over a page thumbnail for a popup display. If the resolution is 
significantly above or below 300 dpi, recognition is likely to 
suffer.

◆ Make sure the correct document languages are selected in the 

OCR panel of the Options dialog box. Only languages included 
in the document should be selected.

◆ Turn IntelliTrain on and make some proofing corrections. This 

is most likely to help with stylized fonts or uniformly degraded 
documents. If IntelliTrain was running, try turning it off – on 
some types of degraded documents it may not be able to help.

◆ Do some manual training, or edit existing training to remove 

unsuccessful training.