Download PDFOpen PDF in browser

Radar-Based Passive Step Counter and Its Comparison with a Wrist-Worn Physical Activity Tracker

EasyChair Preprint no. 9079

12 pagesDate: October 24, 2022


Inspired by novel applications of radio-frequency sensing in healthcare, smart homes, rehabilitation, and augmented reality, we present an FMCW radar-based passive step counter. If a person walks or performs other activities, the individual body segments, such as head, torso, legs, arms, and feet, move at different radial speeds. Owing to the Doppler effect, the individual body segments in motion cause distinct Doppler shifts that can be used to recognize and analyze the performed activities. We compute the time-variant Doppler spectrogram of a walking activity of a person and extract the high energy Doppler components that mainly describe the torso movements during walking. From the computed Doppler spectrogram, we then compute the mean Doppler shift. To detect and count steps, we apply the peak detection algorithm to the mean Doppler shift. Our approach is evaluated using a walking activity data set. We have used a ground truth and a commercially available wrist-worn human activity tracker to validate the results of our approach. Our results show that our system is capable of passively counting the number of steps with an overall accuracy of 98.51% within a 12 m range. Therefore, our proposed system can be used as a passive step counter in indoor environments. Besides, it can also contribute to indoor localization and human tracking applications.

Keyphrases: FMCW radar, Mean Doppler shift, peak detection, Spectrogram, step counting

BibTeX entry
BibTeX does not have the right entry for preprints. This is a hack for producing the correct reference:
  author = {Muhammad Muaaz and Sahil Waqar and Matthias Pätzold},
  title = {Radar-Based Passive Step Counter and Its Comparison with a Wrist-Worn Physical Activity Tracker},
  howpublished = {EasyChair Preprint no. 9079},

  year = {EasyChair, 2022}}
Download PDFOpen PDF in browser