Kofax INDICIUS 6.0 - Manual (Page 53)

Configuring Recognition and Correction
Getting Started Guide (Fixed-Form) 47
X To remove noise from the field
Recognition uses image processing algorithms to strip out unwanted noise from a
document image prior to recognition. Features can include (but are not limited to):
ƒ Dots/speckles
ƒ Lines/Boxes
ƒ Images/logos
Recognition uses knowledge of the dimensions and number of pixels for each feature
on the image, as well as the minimum separation between neighboring features to
deduce which features are noise. Noise removal techniques can be set differently for
different fields for maximum flexibility.
All noise removal properties are defined on the Background tab. Blob removal
provides five default rules for removing marks, alternatively a new rule can be
specified in the window opened by clicking .
X To create a single line field with box removal
Recognition uses image processing algorithms to strip out pre-printed boxes from
character and mark fields prior to recognizing the characters or marks. This is
particularly useful for processing legacy forms where the boxes are not printed in a
drop out color, for example, the character boxes in the credit card number field on
the example image.
1
Scroll the image to view the credit card number.
2
Create a new single line field with the properties shown in Table 5-7.
Figure 5-12. Single Line Field with Box Removal
Table 5-7. Single Line Field Configuration
Property Value
Name: CreditCardNumber
Left: 1012
Top: 2377