Incomplete motion data reconstruction for character animation

Authors: Xiao, Z., Nait-Cherif, H. and Zhang, J.J.

Journal: Proceedings of the IADIS International Conference Computer Graphics, Visualization, Computer Vision and Image Processing 2009, CGVCVIP 2009. Part of the IADIS MCCSIS 2009

Pages: 89-98

Abstract:

This paper explores the technique that employs Principal Component Analysis (PCA) for motion synthesis and incomplete motion data reconstruction. A low dimensional space of human motion patterns are generated from a large amount motion capture data using PCA. The generated motion pattern is then applied to synthesize different scaled human character animations with a different speed by interpolation and extrapolation from existing motion patterns. According to animators' desire, we also construct incomplete motion data for a long and continuous motion sequences. This technique provides flexibility for animators to efficiently re-use existing motion capture data for character animation. © 2009 IADIS.

Source: Scopus

Preferred by: Zhidong Xiao and Jian Jun Zhang