track_cluster(markers, /, cluster, *, include_lcs=False, lcs_name='LCS')[source]#

Fit a cluster to a TimeSeries of point trajectories.

This function fits a cluster to a TimeSeries and reconstructs a solidified version of all the points defined in this cluster.

  • markers (TimeSeries) – A TimeSeries that contains point trajectories as Nx4 arrays.

  • cluster (dict[str, numpy.ndarray]) – A cluster definition as returned by ktk.kinematics.create_cluster().

  • include_lcs (bool) – Optional. If True, return an additional entry in the output TimeSeries, that is the Nx4x4 frame series corresponding to the tracked cluster’s local coordinate system. The default is False.

  • lcs_name (str) – Optional. Name of the TimeSeries data entry for the tracked local coordinate system. The default is “LCS”.


A TimeSeries with the trajectories of all cluster points.

Return type