Substitute common OCR misreads
Use the procedure in this topic to add, delete, or modify substitutes for common OCR misreads.
It is not uncommon for a recognition engine to confuse similar characters such as the letter "O" and the number "0" or the letter "I" and the number "1." Especially for amounts, it is useful to have Tungsten TotalAgility automatically compensate for such discrepancies. To do this, you can define a list of predefined recognition substitutions to automatically correct misread characters.
For a format locator, you can select the Use OCR substitution setting so the defined substitutions are used for fields that have this locator assigned. This list is used globally in the project.
Procedure
-
On the
Project tab, in the
Configuration group, select
Project Settings
.
- Select the Recognition tab to view the recognition settings.
-
Select
OCR Substitution.
The OCR Substitution window and a list of defined substitutions is displayed.
-
Modify the list as follows:
- To add a new substitution, enter the characters into the fields of If OCR reads this and Replace it with, and select Add.
- To delete an OCR substitution, select an item from the list of available OCR substitutions, and select Delete in that row.
- Optional. Select OK to close the Project Settings window.
- Save the changes to your project.