Just to expand, what you describe is one of the branches of AR, which is image/object triggering. Meaning that when the device (smartphone) identifies the image/object an action is triggered. This action could be displaying a video, a 3D object, linking to a site, etc.
However, there's another type or branch of AR, which is geo enabled. My view is that the big part of AR will be that way, that's why we build Terra Icons
Wearables (HoloLens, Meta, something like Google Glass) will enable AR even more.