Skip to main content
. 2021 Feb 19;12:628728. doi: 10.3389/fpsyg.2021.628728

TABLE 1.

List of functions available in OpenPoseR.

Function Description
acceleration_x Computes acceleration for points on x-axis for a given data frame in OpenPoseR format.
acceleration_y Computes acceleration for points on y-axis for a given data frame in OpenPoseR format.
clean_data Discards points with probabilities below a pre-defined cut-off as well as zero values by imposing data.
create_csv Creates a CSV file in OpenPoseR format from raw OpenPose JSON output.
det_hand Determines whether the left, right, or both arms were moved by the actor.
en_acceleration Computes Euclidean norm of sums of acceleration vectors (x,y).
en_velocity Computes Euclidean norm of sums of velocity vectors (x,y).
file_acceleration Computes acceleration for a CSV file generated by create_csv(), using acceleration_x() and acceleration_y().
file_clean Cleans a CSV file generated by create_csv() using clean_data().
file_en_acceleration Computes en_acceleration() for given CSV files generated using file_acceleration().
file_en_velocity Computes en_velocity() for given CSV files generated using file_velocity().
file_velocity Computes velocity for CSV file generated by create_csv(), using velocity_x() and velocity_y().
file_video_index Creates a video index using video_index() for a given directory and save it as a CSV file.
plot_frontal Plot a heatmap of points extracted from OpenPose data using create_csv().
plot_timeseries Plot a time series (velocity or acceleration), derived from OpenPose data using OpenPoseR.
velocity_x Compute velocity for points on x-axis for given data frame in OpenPoseR format.
velocity_y Compute velocity for points on y-axis for given data frame in OpenPoseR format.
video_index Create an index of video files and their properties (length, frame rate, etc.) in a given directory.

Notice that for all of the software’s core functionality, so-called wrapper functions with the prefix “file_” make it possible to directly work with files as input and output arguments, thereby making it easier to handle data derived from large sets of video clips.