I experienced, that when using an image target it is having some offset time by time and is realigning as soon as the image is detected again.
Using the spatial anchor instead is much more stable.
From a technical perspective, that most not be because you could just register the scene with an image and then fixing it with AR functionality of the device as if it works with the spatial anchor.
Is the image target functionality still developed?
How could I improve my experience by using the image target because it is needed to set the orientation of the 3D models?
I use the latest Vuforia Studio, Vuforia View App, iPad 6. Generation, Target Image has a Vuforia Rating 5 out of 5 Stars.
Do you have Extended Tracking and Persist Map enabled for your project? You will find these options on the 3D Container properties:
If not, try turning them on and see if that improves tracking when the image is not in view.
