mocap.averaging: This function averages a list of motion capture recordings.

In browarsoftware/RMoCap: Package for processing and kinematic analyzing motion capture data.

Description

Averaging is performed on each rotation channel of hierarchical model separately with Dynamic Time Warping barycentre averaging (DBA), see: François Petitjean, Alain Ketterlin, PierreGançarski, "A global averaging method for dynamic time warping, with applications to clustering", Pattern Recognition, Volume 44, Issue 3, March 2011, Pages 678-693, https://doi.org/10.1016/j.patcog.2010.09.013 Each rotation signal holds rotation represented as quaternion. Quaternion averaging is performed with Quaternion Markley averaging algorithms, see: . F. Landis Markley, Yang Cheng, John Lucas Crassidis, and Yaakov Oshman. "Averaging Quaternions", Journal of Guidance, Control, and Dynamics, Vol. 30, No. 4 (2007), pp. 1193-1197. https://doi.org/10.2514/1.28949 Results are smoothed with Weighted Quaternion Markley averaging algorithms using Gaussian kernel.

Usage

mocap.averaging(myList, DBAIterationsCount = 50, eps = 1e-04,
  plot.me = TRUE)

Arguments

myList	list of mocap data frames. Algorithm uses columns with names that has .Rx, .Ry and .Rz names. Rotation should be represented by Euler angles in degrees.
DBAIterationsCount	maximal number of iterations of DBA algorithm (default value is DBAIterationsCount = 50).
eps	threshold value for DBA - iteration stops when absolute value of difference between normalized DTW distances on this and previous iteration is less than eps (default value is eps = 0.0001).
plot.me	if TRUE, plots DTW distances for each averaged signal (default value is plot.me = 50).

Value

return object of class averaged.mocap.

Examples

 #load list of objects of mocap class
 data("mawashi.geri.right.list")
 myList <- list()
 #assign data frames to list
 for (a in 1:length(mawashi.geri.right.list))
 {
   myList[[a]] <-mawashi.geri.right.list[[a]]$data.frame
 }
 #run compiled version of mocap.averaging function
 res.data <- mocap.averagingCmp(myList, 50, eps = 0.000001)
 plot(res.data)
 #write results to disc as bvh file
 skel <- set.data.frame(mawashi.geri.right.list[[1]], res.data$fullData)
 write.bvh(path = "avg.mawashi.geri.right.bvh", skeleton.helper = skel)

browarsoftware/RMoCap documentation built on May 16, 2019, 7:28 a.m.