Each camera of the setup should have a minimal overlap (enough to let a person be seen at the same time by different cameras).
Each camera should see one and only one person at a time.
I saw this phrase in sdk overview. Does this mean that I have to calibrate using only one person when I calibrate? If I complete calibrate using only one person, will multibody tracking be applied based on the data?
Hi @hyunsoo
Yes, the calibration must be performed with a single person moving in the field of view of the cameras.
The position of the person’s joints and limbs will be used to calculate the position and orientation of each camera with respect to the others.
This process will not affect the “multi-person” detection capabilities.