To search, Click below search items.


All Published Papers Search Service


Motion Editing Based on Joint Classification and Discrete Blending


Hyoungseok B. Kim, Hosook Kim


Vol. 7  No. 2  pp. 55-60


In this paper, we present a new method of motion editing based on joint classification and discrete blending in order to efficiently reuse captured motion data. First of all, we define a joint energy of a character in order to measure the similarity of motions, and then apply the clustering technology to the set of joint energy. Hence, we may have two groups of joints: major joint group and minor joint group. If the major groups of two given motions have the same cardinality and the error of their motion energy is within the given threshold, then we classify the motions as similar motions. Otherwise, we consider the given two motions different. Our method may enrich the database of motion capture data, so plays an important role in digital contents such as computer game and computer animation.


Motion Editing, Joint-Classification, Motion Synthesis, Blending.