Panasonic IMAGECHECKER PV230 Manual page 30

Segment characters and assign a label for each
character.
Place an object under the camera to let characters in the inspection area.
1.
Select "Register".
The setting window of character image is displayed.
2.
Select "Segment: Execute", or press the F2 key to
execute segmentation.
Characters are cut out.
The segmentation condition at this time is the same as the
condition specified in "2.4.1 Image Filter" and "2.4.2
Segmentation Condition".
3.
Select "Character Assignment" to assign provided labels
to the segmented characters.
A window to input a label is displayed.
Specify labels in order of numbers of segmented characters.
The border for the character currently the label is being
specified is displayed in blue.
Input "_" for the label not to register a part of the segmented
characters. In the right example, only characters 0 to 5 are
registered, and characters 6 to 9 are not registered.
Use the "->" or "<-" button to move to the assignment of a
next character or a previous character.
It is not necessary to register all of the segmented characters
in this step. For details, refer to page 30 "Point of images to
be used for dictionary".Point of images to be used for
dictionary.
Select "Enter" on the software keyboard at the last.
4.
Select "Register: Execute".
A message showing the result of registration appears.
"Registration is done. Execute learning."
It indicates that all the character patterns assigned to the
labels have been registered properly.
"(n) character(s) could not be registered"
It indicates that the character patterns could not be registered
for the specified label as 3 character patterns in the label
have been already registered.
The character images with other labels have been registered.
.
See the next section "Check the characters registered in (Dictionary)" for the details of checking the
character patterns registered in each label.
5. 2 4
Repeat to to register all the necessary character patterns. Press the CANCEL key at the last.
It is recommended to cut out characters with the same threshold method as the character patterns already
registered when you want to add characters after the execution of learning.
28