rate detection

Discussion in 'XeThru Explorer' started by Andrew Walsh, Oct 19, 2016.

  1. Andrew Walsh

    Andrew Walsh New Member

    Just wondering if you can tell us what method you use for rate detection.
    Thinking of writing my own but do not want to reinvent the wheel.
    I am not looking for details, just general statements.

    eg. ECG based heart rate monitors try and pick the peaks then filter the RR interval.
  2. Magnus Bache

    Magnus Bache New Member Staff Member

    Hi Andrew,

    To monitor respiration rate we use a sliding FFT window and look at the frequency peaks in the resulting spectrum. The periodic chest movement of a breathing person will produce a symmetric frequency spectrum, i.e you will have peaks at both the postive and the negative respiration frequency, in addition to higher harmonics. We also have criteria for how "pure" the spectrum is, where we apply a set of frequency masks. The window length is determined by how low respiration rates you want to detect. The longer the window, the lower the frequency.

    Hopefully this was helpful!

    Magnus Bache,
    SW Development Engineer.
  3. Andrew Walsh

    Andrew Walsh New Member

    Yes thanks for that, FFT better than peak picking ...
  4. Nir Regev

    Nir Regev New Member

    Hi Andrew
    You will still need to find peaks in the spectrum. Note that there will be multiple integers of the first breathing harmonic (micro-Doppler effect).
    Take a look:

    Last edited: Nov 1, 2016
    Maria likes this.