Hi, I'm using wikitude 9.1.0.
Unity version I'm using is 2019.3.10f1
My problem is that I have an app that must unlock certain features by scanning a 3D object, in this case is a teddy bear.
The point is that I'm able to track the teddy bear by using a picture instead of actually recognizing the toy.
The point of the app is to get the real toy to unlock the feature, so the object recognition / object tracking is not working for me because it can be replaced with image tracking.
This happens on android devices, not tested in iOS yet.
Anybody with the same issue?
Could you please share the pictures that you used to create the Object Tracker of the Teddy Bear - also did you check the point cloud of the bear within Wikitude Studio? This is a good first indication if the object is (or the input files) are suitable for the Object Recognition. If the point cloud is not very dense, this could be an indication that the recognition / tracking might not be very suitable.
Movable parts (e.g. arms, legs) and the fact that the object itself might not really be well textured (assume it's a teddy with one color) do also have an impact in the creation of the Object Tracker.
We just released an update of our SDK which includes a beta test phase for our 3D Model based Object Recognition feature, which extends the amount of objects also to less textured ones. You can find details on this release here:
Thx and greetings