FOSS
/
BangleApps
mirror of https://github.com/espruino/BangleApps
1
0
Fork 1
Code Issues Packages Projects Releases Wiki Activity
BangleApps/apps/hrmmar
History
Erik Andresen 77e270d356 hrmmar: Add fft elimination algorithm to remove MA 
The fft with size=256 is calculated over a window of 8 seconds with step size 2
seconds. Sample rate of acceleration (12.5Hz) and PPG sensor (25Hz) is unchanged, so the algorithm runs with 12.5Hz.
Generally the maximum peak in fft spectrum in an interval of -5..+10BPM of
last one is used.

The low fft resolution limits the accuracy to ~3BPM, but mean error
is ~5-10BPM.

If firmware reports confidence >= 90% its value is used.
 		2022-11-27 19:05:40 +01:00
..
README.md hrmmar: Add fft elimination algorithm to remove MA 2022-11-27 19:05:40 +01:00
app.png hrmmar: Add fft elimination algorithm to remove MA 2022-11-27 19:05:40 +01:00
boot.js hrmmar: Add fft elimination algorithm to remove MA 2022-11-27 19:05:40 +01:00
fftelim.js hrmmar: Add fft elimination algorithm to remove MA 2022-11-27 19:05:40 +01:00
metadata.json hrmmar: Add fft elimination algorithm to remove MA 2022-11-27 19:05:40 +01:00
settings.js hrmmar: Add fft elimination algorithm to remove MA 2022-11-27 19:05:40 +01:00

README.md

HRM Motion Artifacts removal

Measurements from the build in PPG-Sensor (Photoplethysmograph) is sensitive to motion and can be corrupted with Motion Artifacts (MA). This module allows to remove these.

Settings

  • MA removal

Select the algorithm to Remove Motion artifacts:

  • None: (default) No Motion Artifact removal.
  • fft elim: (experimental) Remove Motion Artifacts by cutting out the frequencies from the HRM frequency spectrum that are noisy in acceleration spectrum. Under motion this can report a heart rate that is closer to the real one but will fail if motion frequency and heart rate overlap.