If I want to display a cube depending on the current pose, can I use scaledCamToWorld() directly? It does not seem to give me the right projection. I do:
With little rotation it seems that just the scale is a little off. With 45deg rotation it seems it is rotating of 90deg. In general it doesn't seem to stick to a specific position in the 3D world, while if I print the depth maps from tracking on the live image they are perfectly following edges. Why is that? Do I need to multiply the pose by the scale? Any clue @JakobEngel?
If I want to display a cube depending on the current pose, can I use
scaledCamToWorld()
directly? It does not seem to give me the right projection. I do:With little rotation it seems that just the scale is a little off. With 45deg rotation it seems it is rotating of 90deg. In general it doesn't seem to stick to a specific position in the 3D world, while if I print the depth maps from tracking on the live image they are perfectly following edges. Why is that? Do I need to multiply the pose by the scale? Any clue @JakobEngel?