4.7.9 Feature Extraction
What Is Feature Extraction?
Feature Extraction binarizes an image captured by a
camera, detects a specified size of clusters of white or
black pixels (hereinafter called "object"), and then judges
as OK or NG depending on the measured number. The
checker outputs the feature quantity such as the
coordinates of a center of gravity or principal axis angle
(+/-90 degrees).
Output Data
Feature quantity of an object
Area:
Gravity X/ Gravity Y:
Perimeter:
Principal axis angle:
Projection
Width/Height:
No. of Objects:
Judgement Results:
Specifying Checker No.
1.
Select "Feature Extraction" in "Checker Type".
2.
Select the checker No. list and specify a checker
number to set.
To select the number not displayed in the list, move the
cursor to display the desired number or specify the
number in "Checker No." field above the table.
3.
Press the ENTER key to determine the checker No.
"Feature Extraction" setting window is displayed. The binary image is displayed on the screen window.
Area of an object (in pixels)
Gravity center X and Y of a detected object
Perimeter length of an object
The angle between the principal
axis angle and X axis
The width and height of the
rectangle circumscribing around
an object
The number of detected objects
If the number of detected objects is within the allowable range you have set, the
object will be judged as "OK", if it exceeds the allowable range, the object will be
judged as "NG".
115