LMI Technologies Gocator 2130, Руководство пользователя, Страница: 161 / 652

Gocator Line Profile Sensors: User Manual
Gocator Web Interface • 161
Using Edge Detection
When using edge detection for part matching, the Gocator compares a model that you must create from
a previous scan to a "target" (one of the parts you want to match to the model).
In the data viewer, a model is represented as a yellow outline. The target is represented as a blue outline.
If the part match quality above a minimum user-defined level, any measurements configured on the
Measure page are applied.
Model (yellow outline) and target (blue outline).
Part match quality is 87.789%, which is greater
than the minimum set by the user, so the parts match.
When you create a model, the Gocator runs an edge detection algorithm on either the heightmap or
intensity image of a scanned part. The resulting model is made up of the detected edge points. The scan
used to create the model should be of a reference (or "golden") part to which all other parts will be
compared.
After the model has been created, you optionally modify the model by adjusting the sensitivity (how
many edge points are detected), or selectively remove edge points from the model, to improve
matching.
Models are saved as part of a job.
Once you have finished modifying the model, you can also modify target sensitivity, which controls how
many edge points are detected on the subsequently scanned targets that will be compared to the
model; the same edge detection algorithm used for creating models is used to compare a model to a
part.
Typically, setting up edge detection to perform part matching involves the following steps: