Hello Apple devs,
I have been making an app with Unity to view houses in AR and I am using an image target to instantiate and position the model on site. I am wondering if the mesh that the device generates during an AR session can be saved and then used as a reference for the next AR session. So, with each AR experience the reference model would grow and grow. In theory, after the first AR session on site, the image target would be unnecessary anymore and the device would localize itself by matching what it sees with the meshes from previous scans. This is a similar concept to what Vuforia is doing with area targets except the runtime scans would constantly be being added to the reference model with each AR session.
Hope that makes some sense. Here is a video of us doing an AR walkthrough to provide context. Thanks in advance.
https://www.youtube.com/watch?v=kzeVVXaKUzM
You are looking for ARWorldMap
or this sample code from 3 years ago.