Tracking, Markers, AR: ArUco?

vvvv beta feature
1

Hello

I’m trying to track AR markers from a camera image. I cannot use the freeframe nodes since I’m also using Kinect2 nodes in the project, and I cannot run two instances of vvvv for performance reasons, loosing time sending data from one to the other.

DetectObject is too jumpy and unreliable.

Does anyone have experience with ArUco in vvvv?
http://docs.opencv.org/3.1.0/d5/dae/tutorial_aruco_detection.html

I’ll try writing nodes for that, I’ll report here how that goes.

Other suggestions/libraries/contributions?

thanks
dominikkoller

2

isn’t there an opencv (and therefor emgucv) implementation of ArUco? Or, isn’t this coming from opencv originally? If that’s the case, i’d consider using the imagepack as basis for the nodes, since you could ‘just plop an example into a node’ and then benefit from the imagepacks threadhandling.

3

ArUco builds up upon openCV, yes. That doesn’t really help though, does it?

4

What exactly do you mean by that?

5

hehe, i mean by that, if you find a good example of ArUco in emguCV, it would probably be quite easy to transfer that code into an imagepack node. at least, that was my experience with that node.

6

it seems a consortium has taken up again work on ARToolkit:

there is also a unity module available - could be a starting point for a port but i guess this is not a quick solution…

7

Ah you’re right, there is Emgu.CV.Aruco, that could be really interesting to add to the imagepack.

8

this looks like a starting point

9

I’ll look into it. Seems interesting.

10

This topic was automatically closed 365 days after the last reply. New replies are no longer allowed.