modifying the camera image on the dashboard

[iblis432]

I was successful in integrating the vision code from the example into the robot code and all works well on that end, but I am having trouble supplying that data to the dashboard. I am trying to send the image of that black and red screen that appears when looking at the rectangle, along with the overlay. My preference would be to have an unchanged image next to the modified image.

My first try was to simply send the image through "Set User Data High" but it returned a blank image on the dashboard.

Next I tried copying the code from the vision processing vi into my dashboard code, hoping I could just have the computer process the square locations and build the new image, but again the image remained blank.

I can post the code this evening. Does anyone know how I can get what I am looking for?

[Alan Anderson]

Originally Posted by iblis432

My first try was to simply send the image through "Set User Data High" but it returned a blank image on the dashboard.

Next I tried copying the code from the vision processing vi into my dashboard code, hoping I could just have the computer process the square locations and build the new image, but again the image remained blank.

Is it a blank white, or a blank black? The red/black processed version is a binary image, and if you try to display that without selecting the binary palette for the indicator it will be presented as black and slightly less black. If that's your problem, you can fix it by changing the palette for the image indicator.

[iblis432]

Thanks for the help, that was not the issue but I got it working now, although honestly I don't know what's all that different.

[PSHRobotics]

Could you provide information about how you fixed your problem?
We are able to see the processed image when we run the dashboard out of labview (by pressing the arrow in our edited Dashboard Main vi). However, when we build the Dashbaord.exe we get a plain/gray screen which to us seems like no data is being sent to the Dashboard when we run it as the .exe

[iblis432]

Why certainly, here is a brief description, and either this evening or tomorrow I will upload the code.

Basically I changed what we wanted. I no longer have the two images combined, and instead have the particle overlay going directly onto the camera image.

I copied the part of the vision processing code that creates the overlay (including the scores limiter, this is important), and I pasted it into the dashboard project. I then took the pre-existant vision indicator and connected it to where the overlay's output is. I'm sure that sounds confusing, so I will be sure to give you a copy of the code ASAP

[PSHRobotics]

Originally Posted by iblis432

Why certainly, here is a brief description, and either this evening or tomorrow I will upload the code.

Basically I changed what we wanted. I no longer have the two images combined, and instead have the particle overlay going directly onto the camera image.

I copied the part of the vision processing code that creates the overlay (including the scores limiter, this is important), and I pasted it into the dashboard project. I then took the pre-existant vision indicator and connected it to where the overlay's output is. I'm sure that sounds confusing, so I will be sure to give you a copy of the code ASAP

Actually it makes perfect sense and that is what we were thinking about doing if we could not get a way to see the other image off the dashboard or if that used too much of the crio's processing. Thanks!

[iblis432]

yep : ) basically the image created on the dashboard is not generated by the cRIO, but instead by your computer, having them work in parallel to compute what to do.

[CRLS97]

Originally Posted by iblis432

yep : ) basically the image created on the dashboard is not generated by the cRIO, but instead by your computer, having them work in parallel to compute what to do.

after having your computer process it, is there a way of sending the data back to the crio so the code running on the crio can use that information to do whatever one programs it to do?

[iblis432]

I do not doubt that there is a way, I just personally do not know how. I have it so both the robot and the computer are doing the same process, I just don't have the lag time of sending processed images to the dashboard

[Alan Anderson]

Originally Posted by CRLS97

after having your computer process it, is there a way of sending the data back to the crio so the code running on the crio can use that information to do whatever one programs it to do?

Here's an example.

[Ross3098]

We are having a lot of trouble with the dashboard processing as well. We are using THE EXACT SAME CODE as the robot vision processing (Rectangular Target Processing.vi) is and it works on the cRIO with no issues (other than massive lag to other robot systems). We wire the unfiltered feed to the dashboard and it clearly shows in the window.

Our problem is with the IMAQColorThreshold vi. The image leaving the vi is absolutely white. Our parameters are exactly the same as the cRIO processing code and we have no other ideas.

[iblis432]

In the attached files are both our robot code and our dashboard code. I hope this helps everyone, and sorry for being slow about getting it to you.

[iblis432]

well then, I'm assuming this helped everyone and answered their questions?