Background - Smartphone-Based Human Activity Recognition - page 13

have also brought new challenges: preserving battery life and minimizing obtrusive-
ness while being able to gather reliable context information from limited sensing.
These sensors are sometimes uncomfortable for the common user (e.g. if they are
fastened too tight or wired or if they need to be constantly repositioned after dressing)
and cannot provide a long-term solution for activity monitoring without recharging
them regularly.
In addition, hybrid sensing approaches, which combinewearable and ambient sen-
sors from different sources, offer an alternative robust option for HAR. For instance,
in Tapia et al. ( 2006 ), a sensor rich environment has been set for the collection of
signals from 72 environmental and body sensors aiming to evaluate complex activ-
ities in an indoor location. In this work, we employ accelerometers and gyroscopes
for the sensing human body motion. We describe their key features as follows.

2.3.2.1 Accelerometer

The accelerometer is an instrument that measures the experienced physical accelera-
tion of an object. It has been employed for several applications in science, medicine,
engineering and industry such as for measuring vibrations in machinery, acceleration
in high-speed vehicles and moving loads on bridges. For what concerns to HAR, the
accelerometer is one of the most widely used sensors for reading body motion signals
(Mannini and Sabatini 2010 ).
Its principle of operation generally consists of a seismic mass which is displaced
in relation to the acceleration it is subjected to. The displacement can then be trans-
duced into a measurable electrical signal. This phenomenon has been applied for
the development of Microelectromechanical Systems (MEMS) sensors. Their tech-
nology allows to create nano-scale devices fabricated with semiconductors. They
are advantageous against other sensor technologies because it is possible to pro-
duce them in large scale and with low manufacturing costs. Most common MEMS
accelerometers work as a capacitive sensor composed of a cantilever beam with a
proof mass whose deflection is correlated with the acceleration experienced by the
sensor (Woodman 2007 ; Yang and Hsu 2010 ).
Acceleration magnitude and direction can be measured as a vector quantity by
orthogonally arranging sensors in the three spatial dimensions. This can be also built
on a single chip and it is now common to find triaxial accelerometers in several
commercial electronic devices. This is the case of smartphones which we exploit in
this research.
One of the problems of using accelerometers for detection of body motion is the
effect of the gravitational field which is always present in the measurements and its
magnitude ( g = s 2 ) is relatively high. However, it can also be separated from

body motion by filtering (see Sect. 4.3.2 ) . Sensing the gravity vector can also help to
determine orientation of an object with respect to the gravitational-force axis when
triaxial accelerometers are used.

9 . 81m /