PCA Analysis of motion data
Principal component analysis  is used we represent motion data as a matrix, we can do SVD (singular value decomposition) and obtain a set of singular values, whose magnitude indicates how important is the particular component in the whole matrix.
The motion capture data are composed of a set of poses over time. The poses can be represented as a 3D position of markers. PCA itself is just one stage in the whole process, It is followed by low-pass filtering or smoothing and followed by polynomial regression.
Smoothing gets rid of the high-frequency components in motion, thus driving less important principal components almost to zero.
- Transforms the Euclideam space of the data into EigenSpace.
- Compute mean vector and covariance matrix.
- Create new PCA on linearized quaternions object.
- Performing eigen analysis
External Links :
Source On GitHub (Coming Soon)