In this paper, we propose the model-free inverse kinematics method based on high order moment feature and its Jacobian matrix, which can use arbitrary shape information as reference input free from the structure of targeted kinematic chains. The method is especially useful to generate the motion of humanoid systems, and we can generate the whole body pose from the set of 3D markers or the pixels 2D images, without labeling each feature point of these data with the specified link of the humanoid. The moment feature can be computed from various types of quantities, for example, from geometric points, mass density, pixel images, and the labels of some featured links, which implies general versatility and applicability of the method. We show some results of motion of a human figure generated by the label-less 3D markers and the 2D images.