Hardware manual

Circular Pattern Find Impact Reference Guide
Datalogic Automation Inc. 3-74
NOTE: This tool has been deprecated and replaced by the Pinpoint Pattern Find tool (see page 3-84). It is
documented here only for existing vision programs and is no longer supported.
The Correlation Pattern Find tool is a general-purpose correlation based pattern find tool. It finds one or
more instances of a trained model image in a user-defined search region. It can find the model at any angle
of rotation.
Input Name What it is
Input Image The image being inspected
Tool Origin The tool’s origin relative to the entire image space
Train Region The region of the image used to train the model. If no area is defined, the
entire image is processed.
Find Shape List The region of the image to search for the trained model. If no shape is
defined, the entire image is processed.
Model Reference
Origin Placement
Where the tool should place the reference origin: in the center or upper
left corner of the Train Region, or use the Model Reference Origin value
Model Reference
Origin
The model reference origin for Use Model Reference Origin Placement
type
Angle Range The maximum angle, in degrees, that the tool searches in the image for
a match of the trained model. The parameters are minus, nominal, plus,
include end, and include start.
Angle Range Pat-
tern
The maximum possible range of rotation of the image in relation to the
trained model
Full 360 Degree Range:
Angle Range Only:
Angle Range and180 Rotation:
Angle Range and all 90 Degree Rotations:
Minimum Match
Score
Matches with a match score larger than this (0 to 100) are reported,
depending on the Find All Matches and Maximum Number of Matches
settings
Match Sensitivity Indicates amount of noise in the image when attempting to match the
model. Higher settings allow rougher edges and causes more thorough
and longer searching
Find All Matches If True, all matches with a sufficient match score are reported (Maximum
Number of Matches is ignored).
Maximum Number
of Matches
The maximum number of matches to report, based on the Match Score
Maximum Intensity
Change from Model
This is how much brighter or darker the match can be, compared to the
model. Since the normalized correlation algorithm can theoretically han-
dle any intensity change, it sometimes finds matches in very dark back-
ground regions. The change is measured as "match average intensity" /
"model average intensity", so that a range of 0.5, 2 means that the model
can be from half the intensity of the model up to twice the intensity of the
model, thereby avoiding matches in dark background noise.