Digital Signal Processing Reference
In-Depth Information
3.1
Step Detection with Accelerometers
With step we mean the displacement of one foot during walking movement, i.e. the
distance between two consecutive foot prints. The occurrence of a step can be easily
detected from the signal pattern of the vertical acceleration component [ 20 , 21 ] .
However, this approach is sensitive to orientation errors of the sensor unit, as it
is assumed that one axis is aligned with vertical or that the transformation to the
vertical is known. Other possibility it to compute the magnitude of the measured
acceleration vector, i.e. the norm of acceleration [ 13 ] .
The signal pattern varies according to where the user attaches the sensor unit [ 18 ] .
Typical choices to wear the sensor unit are on the belt, e.g. on the side of the user or
on lower back, or onto upper parts of the torso, e.g. attach it to the shoulder strap of
a backpack or wear it in a chest pocket. Another option is to mount the sensor unit
to the shoe of the pedestrian. With a shoe mounted sensor unit the estimation of the
distance traveled can be based on the analysis of the acceleration waveform [ 27 ] or,
if sensor unit includes high quality gyros, also traditional INS mechanization with
zero-velocity updates can be used [ 7 ] . In this presentation, only PDR systems using
torso mounted sensors are considered in detail.
Step detection is often based on the detection of signal peaks [ 18 ] or crossings
of the signal with its average [ 13 ] or some other reference level [ 21 ] . For example,
step detection from acceleration norm consists of the following steps:
1. Low pass filtering and resampling the signal; sampling frequency in the range
20-25 Hz is high enough.
2. Computation of the norm of current acceleration sample, i.e.,
a x (
a n (
t
)=
t
)+
a y (
t
)+
a z (
t
)
(6)
where a n (
t
)
is the acceleration norm and a x (
t
)
, a y (
t
)
,and a z (
t
)
are the filtered
components of the measured acceleration.
3. Instances of step starts t s (
are detected by observing the g -crossings of the
acceleration norm that are followed by a rise rate and a peak height that exceed
the preset limits, and requiring that the time between the current and previous
g -crossings is long enough.
4. The step end t e (
k
)
is considered to be found when the next step starts or when
a predefined time, considered as the maximum duration of one step, has passed
after the start of the current step.
k
)
An example with acceleration norm and the detected step starts is shown in Fig. 4 .
The data for the figure were recorded using a sensor unit that was attached to the
belt and positioned to the back of the test walker.
 
Search WWH ::




Custom Search