Synthesis of Human Motion Using Kalman Filter
This paper is concerned with the post-processing of motion-captured data. The post-processing is needed for several reasons: jerky motions due to sensor noise, violation of body constraints such as extraneous joint D.O.Fs, generation of new motion by editing existing motion data, and application of a motion to different character models. In this paper, the process of generating animated motion is viewed as a dynamic system, which takes the captured motion as input. Within a single Kalman filter framework, we were able to handle the following problems effiectively: satisfaction of physical constraints inherent to human body, user-specified kinematic constraints, motion transition, and noise reduction.
KeywordsKalman Filter Motion Data Human Motion Kinematic Constraint Motion Transition
Unable to display preview. Download preview PDF.
- Armin Bruderlin and Lance Williams. Motion signal processing. In ACM SIG-GRAPH’ 95, pages 97–104, Los Angeles, CA, August 1995. ACM Press.Google Scholar
- C. Rose, B. Guenter, B. Bodenheimer, and M.F. Cohen. Efficient generation of motion transitions using spacetime constraints. In ACM SIGGRAPH’ 96, pages 147–154, New Orleans, LA, August 1996. ACM Press.Google Scholar
- Andrew Witkin and Zoran Popovic. Motion warping. In ACM SIGGRAPH’ 95, pages 105–108, Los Angeles, CA, August 1995. ACM Press.Google Scholar
- Munetoshi Unuma, Ken Anjyo, and Ryozo Takeuchi. Fourier principles for emotion-based human figure animation. In ACM SIGGRAPH’ 95, pages 91–96, Los Angeles, CA, 1995. ACM Press.Google Scholar
- Michael Gleicher. Motion editing with spacetime constraints. In Symposium on Interactive 3D Graphics, 1997.Google Scholar
- Michael Gleicher. Retargetting motion to new characters. In ACM SIGGRAPH’ 98, pages 33–42. ACM Press, 1998.Google Scholar
- J.C.K. Chou. Quaternion kinematic and dynamic differential equations. IEEE Transactions on Robotics and Automation, 8(1):53–64, February 1992.Google Scholar