Skeletal Methods of Shape Manipulation

The geometric skeleton is derived from a static object using an implicit ‘directions’ method; an IK skeleton is derived from and used to manipulate the geometric skeleton. The model may be reconstructed from the modified skeleton using implicit distance and convolution methods. Introduction We examine the use of the skeleton to manipulate erstwhile static 3D models. These models may be generated by a design system or a hardware scanner, and may be disseminated through the public domain, catalog services, or scanning services. Originally static, the models may be positioned, oriented, and scaled, but not articulated. They may contain separate parts (such as a car and four wheels), permitting only limited animation. In order to animate such a model, we first extract its skeleton. Then we articulate the skeleton, from which we reconstruct the animated surface. The ’IK skeleton’ is traditionally used to control an articulable (i.e., nonstatic) model. More recently it has been applied to static models, but this approach has several problems. The skeleton must be created, usually manually; and a correspondence between skeleton and surface must be established. Articulation is limited and arbitrary metamorphosis is not possible. Geometrically, the ’skeleton’ has the precise meaning of medial axis (or surface) of the object. The medial axis is similar to the IK (inverse-kinematic) skeleton, but is (typically) two-dimensional, whereas the IK skeleton is one-dimensional. The following figure depicts an object, its geometric skeleton, and its IK skeleton. The geometric skeleton consists of 2D surfaces and 1D curves, and is completely surrounded by the object. Every point on the Figure 1. Skeletons and Surface IK Skeleton bone tendon vein muscle