clean_data: Remove points with low probability and points with zero...

In trettenbrein/OpenPoseR: Analyze Motion-Tracking Data Derived from Video Files Using OpenPose

Remove points with low probability and points with zero values from a data frame

Description

Removes points with zero values (i.e. where the tracking of a point failed completely) as well as points with extremely low confidence rating (default: .1) or a given confidence rating (cutoff) from the data frame provided.

By default, for points below the given confidence rating the frame is set to x,y = 0. In a second step, all points with zero values are replaced with mean values (x,y) derived from the previous and following frame.

Usage

clean_data(data, model, cutoff)

Arguments

data	A data frame with separate clumns for every points x, y, and c values (OpenPoseR format).
model	Specify which tracking model generated the data. Supply either body25, hands, or face.
cutoff	Optional. Defaults to .1 as threshold and removes points (x,y) with confidence rating (c) below the cutoff from a frame.

Examples

clean_data(x)

trettenbrein/OpenPoseR documentation built on April 22, 2022, 4:03 a.m.