We’re starting to dig into the 2013 LabVIEW vision example, and my team members had some questions that I’m sure others will have.
Some values on the front panel are not self explanatory, and digging into the code further doesn’t exactly reveal what they are meant to do.
What does the Area Size Limit % represent?
In the score limits, what do “Center” and “Outer” represent?
Good questions. Once the image is thresholded, the result is a binary image that consists of background values and blobs or particles of foreground values.
When processing particles, the Area size limit is used to eliminate all particles smaller than the limit. The units are in % of camera view area. Using a percentage allows it to be independent of camera resolution.
The Center and Outer are terms that refer to the 3 point and 2 point targets. The targets have different aspect ratios, and the numbers are also used to classify whether a given particle is believed to be Center or Outer.
Let me know if this is enough info.
Thanks Greg, that’s great info.
We did some cad work today and realized that from most shooting positions near the pyramid, the standard-issue axis camera with a 47 degree viewing angle will not be able to see all 3 targets at once. So we can use Area size Limit to potentially exclude a partial side target that will have an incorrect center point.
I assume the vision example will also pick up the vertical stripes and treat them as rectangles. Is that correct?
This is one of the reasons for the aspect and hollow tests. It will score well as a rectangular blob, but very poorly on aspect and very poorly on hollow.
Ah, I see. So tightening the aspect ratio will exclude the side targets, but not the top target or the vertical stripes (which both have fairly high aspect ratios).
Very very nice. This does nearly everything teams need right out of the box.