A couple options you can take moving forward.
(1) Make the image target tarcker better
We recommend using an image target with the following characteristics:
- No reflective surfaces
- Varied and abundant detail
- High image contrast
- Absence of repetitive patterns
- Minimal whitespace
- Preferably a photograph or image with high local contrast, as opposed to logos or vector-based graphics
- Avoid using images with excessive text
For further guidance on optimizing image target tracking, please refer to our documentation: 8th Wall
Additionally, you may find this youtube tutorial helpful
(2) You could reference this thread where you could place a 3D model in SLAM space on xrimagefound event