how does this work?

[DesmondLIJIAN]

i do not understand how it works, no need for tags ? how do you know the transformation between kinect and some point rigidly attached to end effector?

[nmssilva]

Hello @RoboDesmend, sorry for such a late reply.
The base level of the jaco arm is given by the yellow paper on the table.
The joints of the jaco arm are detected by the camera using colored elastic bracelets.
Those are basically the "tags" used.
You can see both in this picture:

The jaco gives out its pose so you can calculate the position of the arm's base joint to be in the same level as the yellow paper. After that, you can calculate the relative rotation of the arm using the depth camera data and the jaco's pose.