Robust Kinematic Constraint Detection for Motion Data

Abstract

Motion capture data is now widely available to create realistic character animation. However, it is di?cult to reuse without any additional information. For this reason, annotating motion data with kinematic constraints is a clever step to ease further operations such as blending or motion editing. Unfortunately, prior automatic methods prove to be unreliable for noisy data and/or lack genericity. In this paper, we present a method for detecting kinematic constraints for motion data. It detects when an object (or an end-e?ector) is stationary in space or is rotating around an axis or a point. Our method is fast, generic and may be used on any kind of objects in the scene. Furthermore, it is robust to highly noisy data as we detect and reject aberrant data by using a least median of squares (LMedS) method. We demonstrate the accuracy of our method in various motion editing contexts.