Help with vision assistant tote pictures.

[CrunchyTortilla]

Hi Team 5186 here,
If anyone could provide me with the pictures of a yellow tote that they used to program their vision system in NI Vision Assistant that would be great! The Ideal pictures would be a front view and a right and left view of the tote with the reflective tape reflecting green light coming from a ring light. Any pictures are greatly appreciated!
Many thanks! -5186

[Joe Ross]

Have you looked at the pictures included with the labview vision examples?

[CrunchyTortilla]

No I have not. Does the reflective tape reflect green light? Does it even matter?

[faust1706]

Originally Posted by CrunchyTortilla

No I have not. Does the reflective tape reflect green light? Does it even matter?

Retro-reflective tape redirects light directly at the source, aka it will be whatever color you throw at it (that includes IR!). It matters when it comes to calibrating your vision in the threshold step.

[CrunchyTortilla]

Very cool i did not know that. So could anyone provide me with any pictures?

[Fauge7]

sample pictures are only so good, the problem you run into is the different rgb values for each regional. They all have slightly different values so one somewhere like Vegas is different then the Hawaii one.

[CrunchyTortilla]

Well i'm a little desperate. I have no contact with my team in my current situation so any pictures would be very helpful.

[faust1706]

Are you looking for yellow totes, gray totes, green bins or all 3? I could dump all the color images (about 5k, 640x480) captured by our robot at the st louis regional if you'd like.

[CrunchyTortilla]

All of them would be very helpful! Many thanks!

[CrunchyTortilla]

Anywhere to even begin would be nice.

[thatnameistaken]

The sample images included with LabVIEW should be in the installation directory at:

C:\Program Files\National Instruments\LabVIEW 2014\examples\FRC\roboRIO\Vision\2015 Vision Example\Sample Images

Or for 64-bit Windows:

C:\Program Files (x86)\National Instruments\LabVIEW 2014\examples\FRC\roboRIO\Vision\2015 Vision Example\Sample Images

If they're insufficient, I'll be happy to take pictures in different lighting/angles tomorrow when I have access to our tote.

[Fauge7]

I have a vision program that uses opencv in java, Faust is also really helpful for c++, once I recover my laptop's hard drive I'll upload the pictures that I had of Vegas and a Arizona. What do you need vision for?

[CrunchyTortilla]

Very thankful for the samples but i am unsure on how to access them? Is the path correctly defined?

[thatnameistaken]

Originally Posted by CrunchyTortilla

Very thankful for the samples but i am unsure on how to access them? Is the path correctly defined?

They should be on the hard drive of the computer with LabVIEW installed; simply navigate there on your computer's file browser. If you're not using LabVIEW or don't have access to your team's programming laptop, I can PM you with a link containing a zip of the images.

[faust1706]

Here is the image dump: https://www.dropbox.com/s/1cacosu1wi...Images.7z?dl=0

Sorry about hosting it on dropbox, as well as the delay from my mention of the images to me finally posting them, I was at the lab for the afternonn and just got back. @cmastudios sent me the .7z through email, and I don't feel like setting it up on the team's github page for this year's code. That's a students job

Ignore the depth images, as they won't be of use to you.

If you have any questions, feel free to ask them on here or via a pm.

I wish chief delphi would take on user references, that'd be nice.

This includes a decent amount of junk instances of the program, as in the image is the same, so I apologize.

Here is an individual image: https://www.dropbox.com/s/vycrgi02t1...t_189.jpg?dl=0

this is something you probably don't have to take into account, but the tote is whited out due to the camera's internal parameters (aka white balance), off set due to going from dark to light. If your camera is going to be below your intake, prepare for this to happen.

If you are going to be tracking the reflective tape and throw light at it, then this images aren't much use. This image might be useful if you are developing your own vision system: https://www.dropbox.com/s/ayqj2us4iq3y1ee/ir.png?dl=0

This is an encompassing example output of the thresholding step.