Multiwire Program
© 2020 F&S BONDTEC Semiconductor GmbH. All rights reserved.
44
To ensure an exact manual positioning of the crosshairs by the operator, the adjustment points should be selected
in such a way that the crosshairs run at a clearly visible corner. The model area should be moved so that a structure
actually occurs only once in the search area.
9.10.1 Image Capture
In the tab <Capture> the camera and light settings are defined.
The camera settings must be selected so that the contrast of the structure to be searched for stands out as well as
possible against the background. These settings are different for each substrate, so you can never make a rule of
thumb.
TIP
On some surfaces, the Saturation channel sometimes works better. (such as: gold coatings on white sub-
strate)
9.10.2 Modell Creation
In the tab <Model> you can choose between Contour or Grayscale as method.
While Contour Search looks for edges in the image, Grayscale compares the distribution of the different gray tones
in the selected image section.
With Downscale the image resolution can be adjusted. For coarse structures you can work with a larger downscale
to blur the image, which can be helpful in such a case.
With Countour there are two more parameters. Min Contrast defines the minimum threshold value for a detected
edge. Min Length filters out shorter contour pieces.
TIP
Contours should be clear and continuous, many small short structures that do not really form a continuous
structure are not well suited for a stable automatism. But if there is only one structure in the search window
because it is a lonely adjustment point, it can also work quite well with short lines.
9.10.3 Control Search
In the tab <Search> you can change the minimum search quality and the search range of the contour. In this tab, the
image recognition searches the set structure and indicates possible alternative finds.
TIP
Make sure that only one contour is detected and no alternative finds are made. If this is unavoidable, make
sure that the set <Min Score> is significantly higher than the score achieved with the alternative finds.