about improving targets precision and stability I have already raed those articles
But I would like to go even further.
Which kind of target grants the higher precision? There are differences between Image Target, Cylinder, Multi Target, Object Target, and others?
Can I use 2 or more targets tracked simultaneously to improve precision and stability?
I tried a simple arithmetic mean between every member of the model view projection matrix (I work with Android), the 3D model is shown in the middle of the 2 markers, do you think this can be a way to improve accuracy, to work better when the marker is far or in the border of the video or to have any kind of benefit?