How to use Sensor Streaming Camera Intrinsics in C# (Unity project) about hololensforcv HOT 5 OPEN

@mauronano I see you created a separate issue for this, I've responded there with some details.

from hololensforcv.

In short, you can't. I had to write a custom little DLL to handle that since the depth camera intrinsics object is not the standard one, but a custom interface to an otherwise opaque structure. There's actually a lot of other issues when trying to do this in Unity. I've spent the past four months working on it, and it's still not exactly right. There's a bunch of coordinate transforms that need to be taken into account because the HoloLens uses coordinate systems typically used by roboticists, while Unity uses another one entirely. It makes using the matrices and projection information that much harder. Not to mention, the two samples for the Kinect, that seem like a good reference, aren't actually how the HoloLens data works in the end.

from hololensforcv.

@FracturedShader could you suggest a workflow to align depth frames with the ones from the color camera?( i mean the standard high resolution color camera, not those made accessible along with the depth and IR sensors)
I think first step would be acquire synchronized frames from the cameras, but i'm not sure how to achieve that since they are from different source groups; then get the intrinsics of the depth camera, and i would appreciate some hints about this part too.

from hololensforcv.

@FracturedShader can u refer to the other issue ?

from hololensforcv.

@cyrineee, sure, that was issue #64

from hololensforcv.