MarilynKeller
commented
1 year ago Indeed the skeleton does not have a kinematic tree but a translation and a rotation per bone group. In the paper, we defined landmarks corresponding to the STAR joints to help the DXA fitting. But we don't have skeleton joints locations for say.
The paper mentions landmarks and joints. It seems that the skeleton doesn't have a tree structure or joints don't control the rotation of the bone...? Changing sp.t and sp.r_abs is change the translation and rotation in local space. The rotation center is the mean of the vertices' locations What is the purpose of joints? Align the joints of the skeleton to the joints of the star model using cost calculation?