Skip to main content
Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

< Back to Article

Table 1.

The joints from the Azure Kinect Body Tracking SDK included and excluded in our analyses.

(For reference see [19]).

More »

Table 1 Expand

Fig 1.

Sketch of the setup viewed from top and side.

A: Setup from top view. Mannequin was standing on the treadmill facing the camera. An infrared light was installed in the background. The room’s window is not visible from the mannequin’s position. Next to the window is a glass pane to the room next door. At some positions of the sun, the sunlight is reflected in this pane when the blinds are up. Ceiling lights are not shown. B: Setup viewed from side. Mannequin was standing on a treadmill facing the camera with a light sensor installed on its left upper chest. In the background, an infrared light was installed. Not shown in the figure: the ceiling lights.

More »

Fig 1 Expand

Fig 2.

Mannequin standing on the treadmill in experiment 1 from the camera’s point of view in the SeFallED setup.

More »

Fig 2 Expand

Table 2.

Settings of the lights and blinds for all nine recordings, as well as the mean measured light (in lux) and infrared intensities (unitless).

More »

Table 2 Expand

Table 3.

Minimum (Min), maximum (Max), median with interquartile range [IQR] of the random errors in mm for all light conditions in experiment 1 measured in the three areas belly, right knee, and right ankle with its surrounding area.

More »

Table 3 Expand

Table 4.

Mean Distance Error values (in mm) for experiment 1 for all light conditions and all relevant joints, as well as median with interquartile range [IQR] of the MDE values of each recording.

The lowest MDE per joint are in bold.

More »

Table 4 Expand

Table 5.

Minimum (Min), maximum (Max), median with interquartile range [IQR] of the random errors in mm for all recordings in experiment 2 measured at the three areas belly, right knee, and right ankle with their surrounding areas.

More »

Table 5 Expand

Fig 3.

Sample visualization of random error values in experiment 1.

Random error values for all pixels in the area of the right ankle with its surroundings at light condition LightOff_IrOff in experiment 1.

More »

Fig 3 Expand

Fig 4.

Boxplots for distances between left and right ankles for all light conditions in experiment 1.

More »

Fig 4 Expand

Fig 5.

Peak in positions of light condition LightOn_IrOn for experiment 1.

X-axis (points right of camera’s focal point), y-axis (points down), and z-axis (points forward) of the ANKLE_LEFT for experiment 1 and light condition LightOn_IrOn with a peak between frames 7000 and 8000.

More »

Fig 5 Expand

Fig 6.

Black spot in light condition LightOn_IrOn for of experiment 1.

Detail of the depth images of frames 7762 to 7767 of light condition LightOn_IrOn (experiment 1) with a black spot moving across the mannequin’s legs.

More »

Fig 6 Expand

Fig 7.

Sample visualization of random errors in experiment 2.

Random errors for all pixels in the area of the belly in recording_1 of experiment 2.

More »

Fig 7 Expand

Table 6.

MDE values (in mm) for experiment 2 for all recordings and all relevant joints, as well as median with interquartile range [IQR] of the MDE values of each recording.

The lowest MDE per joint are in bold.

More »

Table 6 Expand

Fig 8.

Boxplots for distances between left and right ankle for all recordings in experiment 2.

More »

Fig 8 Expand