The Vuforia Spatial Toolbox and Vuforia Spatial Edge Server make up a shared research platform for exploring spatial computing as a community. This research platform is not an out of the box production-ready enterprise solution. Please read the MPL 2.0 license before use.
Join the conversations in our discourse forum if you have questions, ideas, want to collaborate, or to just say hi.
Note that you must own 1-6 Microsoft Azure Kinect cameras to use this software.
Install Git and/or Github Desktop. Install Unity 2017.4.10f1 from the Unity download archive. Install Azure Kinect Development Kit v1.3.0 and Azure Kinect Body Tracking SDK v0.9.4
If you encounter the error DllNotFoundException it will be necessary to also
install Visual Studio Code 2017 and 2019 with support for C++ development. This
covers missing dependencies on freshly installed copies of Windows.
Run this repository's script checkout.sh in Git Bash. This generally looks like the following:
cd Documents/GitHub/repos/Virtualizer
./checkout.shThis should copy some dlls from the Azure Kinect SDKs, you should have some new dll files in the same directory as checkout.sh. The whole process should look like this:
Next, it's time to add OpenCVForUnity v2.4.2 to the project. This project is available on its website. For testing purposes the free trial is enough. Add this asset through Unity's Assets > Import package > Custom package menu.
Open the project in Unity. It should look like this:
If the start scene doesn't open by default, open it here:
Clicking the run button should pop up with uncalibrated cameras. To get an initial calibration click the Autocalibrate button then manually refine the camera positions if desired.
Click "Save Calibration" to save this new calibration.
Download and setup the edge server as described in the tutorial. You should have a working edge server with a normal directory structure:
Stop the server if it's been started.
Download the remote operator from its repository and unpack its directory in your edge server's addon directory. It should look like this:
Stop the Unity scene if it's currently running. Start the server with npm start:
With the server running and logging messages to the console, start the Unity project. It should connect to the edge server. Note the 'name function has been called with: station' messages below
Finally, open localhost:8081 and select your local IP from the top left menu to see the Virtualizer in action.
