Vision Example Breaks on Edit

I implemented the LabVIEW Vision Example into my robot code successfully, but when I try to replace the algorithm that calculates height ratio of the two stripes to one that tells if they are close to the same height, the code breaks. A whole chain of low-level VIs in Rectangle and Cull and appear faded and do not execute. Any help getting edits to work would be much appreciated.

Note: I am not given an option to relink to subVI when I right click the faded VIs like i did with the Vision Processing VI.

It may be helpful to us if you tell us what edits you are making specifically. Otherwise we’re just guessing, which wastes everyone’s time.

I am editing the point where it calculates the fifth target score by dividing one height by hte other and then dividing by 2. I am changing that to subtract one height from the other, take the absolute value, divide by 20, subrtact that from 1, and coerce it to the range 0-1.
That is the edit.

This is probably not it but did you copy the dependencies of the Vision VI into your robot project?

Yes, indeed I did. I’ll upload screenshots of the broken code in a few if I can.

Well, whaddya know, it was when I transferred it on a flash drive to another computer that dependencies didn’t copy correctly… or something like that. Thanks :slight_smile:

Here’s a folder with images, this is what comes up in those bad block diagrams. As you can see, low-level VIs are faded, which should not happen. This only happens on my laptop, my 32-bit desktop doesn’t have this problem.

Currently reinstalling LabVIEW, hopefully that fixes the problem. Fingers crossed heavily.

I couldn’t see any faded VIs in the screenshots. Is the run arrow broken? If it is broken, click it and see what it says is wrong.

The run arrow is not broken. If you look in Cull and Group, there is a VI that is clearly faded, the Max&Min VI in the lower right that comes after add array elements.

The Max&Min function shows up as a bit lighter than the other functions for me too; but it still works. As long as the code runs and you’re not getting any other errors I wouldn’t worry about it. I think they just made the color of the lines in it a bit lighter than everything else.

Side note:
Technically, the Max&Min (and any other block that cannot be opened) is a function, not a VI. Basically, it’s as close to binary as you can go in LabVIEW before getting into the source code of LabVIEW itself.

That’s the problem- that part of the code doesn’t run properly.

What do you mean by doesn’t run properly?

If you’re not getting a broken run arrow, I’d assume it’ll run. But, you’re not getting results you want. What would you expect? What are you getting?

Have you tried any of the debugging tools to see what values are coming out? (Probes would be a good choice here). If you probe, do you see the expected value going into and out of that function or is there something unexpected here?

As a heads up, if VIs are missing, you’d expect to see a big grey question mark instead of their icon. That’s usually the best way to determine you have a dependency issue. Instead, it looks like we’ve got to figure out what has changed and how it has the impact you’re seeing.

Probes indicated zeroes where there should have been 40s or so in the score output.

The most obvious workaround to this issue is bringing my desktop to school, and that’s just not an option for obvious reasons.

The VI’s that appear disabled behave just like that-disabled. But instead of putting out “not executed,” they put out zeroes. I’m confused. Maybe it was something else, but I’ll have to wait until I can access the robot to find out.

Are you able to see images anywhere? If the code hasn’t changed other than moving from one PC to another, I’d expect you to have one of two problems:

  1. all of the dependencies didn’t come along with the move
  2. a hardware configuration is causing you heartache

I’m less inclined to see the first as a lost dependency SHOULD break the VI so that you can’t run it.

That leads me to want to look at the hardware. If we can determine what we see with just a generic video feed (for example, to the dashboard), we can see if it makes sense that we’re not working with images and that’s why the score is showing up as 0.

Sorry everyone, I was in error. The grayed functions/VIs appear to be working normally, and the threshold VI in the example is acting up, getting max and min hue values mixed up or something.

Admins/moderators, you can close this thread as it won’t be helpful to anyone really. It’s your call.