I was not sure in which category I should classify the problem regarding the image targets, therefore now here in the general discussion. The following problem: We are currently developing an AR animation for a castle with the help of Vuforia in Unity. A historical person speaks to the audience. But since this person is only allowed to appear within the app and only at certain places of the castle, we use the image targets. These should practically "unlock" the person.
The targets themselves are glued to the wall, currently we use DIN A3/DIN A4 sizes for tracking. The character that speaks to the audience is placed vertically with the back to the picture. The character is, because it is shown in human size, about 4 times as big as the actually tracked picture. There are now two problems:
1. As soon as Vuforia recognizes the target, it tries to "reposition" the object and the character is extremely shaky. My solution idea, but unfortunately I have no idea how to implement it: If the image was recognized, the character would have to be placed once and the positioning would have to be stopped. By the way, this also works well, if the image is out of sight, then there is no more attempt to place the character and it remains smooth in one place.
2 The character itself is assigned an animator and sound in Unity. When I look at the animation in Unity, it is displayed as a whole and complete. But if I look at the AR representation, there are strange corners and cakes on the character, or distortions that don't belong there. What could be the reason?
I'm looking forward to your feedback and many thanks in advance