I have seen numerous AR application behaving like this: there is a dot in the middle of the screen and we can position that dot on some objects and some content is displayed (I attacked an image if i was not clear enough). My question is how is this kind of behaviour obtained in Unity, my guess is that from that point you cast a ray, but I don't think that AR placed objects, from an ADF for example, can be found with the hit from that ray. The dot selecting objects placed on AR
I have made it work, with the aid of the Google Tango's Area Learning demo scene. I've placed some objects in the area and I have started to send a Raycast from the middle of the camera with "ViewportPointToRay" method. When that Ray would collide with a GameObject you can implement whatever functionality you need.