Is there some reason why the Unity sample project would limit the number of tracked skeletons to 2 people? (Please note, for the purpose of testing I am NOT using the PoseUser.cs script, which seems to limit the number of bodies to 2.)
I am seeing additional ‘bodies’ but only 2 of them ever contain a skeleton with tracked joints.
Has anyone else been able to view 6 skeletons? The documentation seems to indicate support for 6 tracked skeletons, but I have not been able to confirm this in my testing.
I am using Unity 2018.2.11f1, with the recently updated AstraPro Unity project v2.0.12.
We are having the same issue where it only full body tracks 2 people. We’re running on unity 2018.3.6f1.
The SDK code (unity and AseraDotNet.dll) also seems to imply that it should be able to track 6 people.
I put the astra in a usb 3.0 port so it should be fine streaming all this data.
There are 2 different categories of tracking. The first is “skeletal tracking” and it can track 2 people. The second type of tracking is “person tracking”, which can track up to 6 people.
“Skeletal tracking” gives you position data for an array of “joints” (aka “nodes”) that together comprise a whole individual.
“Person tracking” just gives you position data on the “center of mass” of each detected person.
Also, for what it’s worth, the cameras update at 30 FPS (maximum), and the tracking library is very CPU-intensive.
I’m aware this is an older thread but is it possible to restrict the Skeletel Tracking to just one user?
We are having an issue using the Unity SDK package of the camera confusing large cardboard boxes as humans so it creates a second user ID and messes up our joint readings…
Sure. You’d just need to edit the sample code to prevent additional skeletons from being processed. However, you might run into a problem where the boxes are detected FIRST, and then the actual person is ignored. So, you might want to look at the skeletal nodes for some threshold of motion, ignoring a non-moving skeleton, instead. Keep in mind that even static object will appear to have SOME motion, due to the (in)accuracy of the depth sensor on a per-frame basis.
Alternatively, you could look at the depth of the skeleton (distance from the camera), and ignore elements that are outside your “play area”. I don’t know if this would work for your project, though.