2 Kinects to Improve Accuracy at Driver Level? RRS feed

  • Question

  • Hi,

    Would be great to be able to have 2 Kinects to improve accuracy without having to do any extra coding.  Would this ever be supported at the driver level, maybe by some type of setting that allows you to combine the data and treat it like one Kinect?



    Thursday, March 20, 2014 6:24 PM


  • The current sensor does not support such a feature. Keep in mind if such an option would be available, it would not be disclosed on a public forum. We have provided some guidelines when dealing with multiple sensors located here:

    There are too many factors to ensure such a system would work. You would require a full systems that has explicit knowledge of the space, the camera location/orientation and how each camera relate to each other. Typically something like this is going to be a fixed specific build, which a general SDK would not be able to accommodate.

    Carmine Sirignano - MSFT

    Friday, March 21, 2014 9:57 PM