Study of Feature Extraction Techniques for Sensor Data Classification

Study of Feature Extraction Techniques for Sensor Data Classification

Anupama Jawale, Ganesh Magar
DOI: 10.4018/IJICTHD.2021010103
OnDemand:
(Individual Articles)
Available
$37.50
No Current Special Offers
TOTAL SAVINGS: $37.50

Abstract

Human activity recognition is a rapidly growing area in healthcare systems. The applications include fall detection, ambiguous activity, dangerous behavior, etc. It has become one of the important requirements for the elderly or neurological disorder patients. The devices included are accelerometer and gyroscope, which generate large amounts of data. Accuracy of classification algorithms for this data is highly dependent upon extraction and selection of data features. This research study has extracted time domain features, based on statistical functions as well as rotational features around three axes. Gyroscope data features are also used to enhance accuracy of accelerometer data. Three popular classification techniques are used to classify the accelerometer dataset into activity categories. Binary classification (run -1 / walk-0) is considered. The results have shown SVM and LDA when used with rotation and gyroscope data gives the highest accuracy of 92.0% whereas FDA shows 84% accuracy.
Article Preview
Top

1. Introduction

Sensors are the devices used to detect events and changes in it’s environment and send them in the form of various signal to a processing device. In this paper inertial sensors (accelerometer and gyroscope) data is used to classify human activity. Sensor is always paired with some electronic processor device. Brief description of sensors used to recognize human movement activities is given below.

1.2. Accelerometer

Accelerometer sensors and basic vibration-based activity sensors use a piezoelectric crystal that produces a small amount of electrical current in response to motion. The piezoelectric vibration activity sensor recognizes vibration from up and down motion, whereas the accelerometer recognizes anterior and posterior motion. The accelerometer is the most widely used form of rate-adaptation sensor because it is simple, easy to apply clinically, and rapid in the onset of rate response.

The raw data of accelerometer involves three values, Acceleration about x-axis, Acceleration about y-axis and acceleration about z-axis. The acceleration is defined as rate of change of velocity of a device with respect to time. Unit of acceleration is the meter /second squared (m/s2). Along with acceleration, three more terms are also used to define rotation about 3 coordinate axes, viz Roll, Pitch and Yaw. They can be define as follows (Arcoverde Neto et al., 2014).

  • 1.

    Rotation around the front-to-back axis is called roll.

  • 2.

    Rotation around the side-to-side axis is called pitch.

  • 3.

    Rotation around the vertical axis is called yaw.

Pitch, yaw and roll are the three dimensions of movement when an object moves through a medium.

For a moving body in 3d-coordinate space, calculating roll, pitch and yaw angles finds the orientation of the 3-axes, tightly attached to the body, with regards to the reference coordinate system. Trying to get the orientation of the body fixed 3-axes plane (Figure 1).

Figure 1.

Defining Roll, Pitch and Yaw with respect to floor for a human body, assuming accelerometer device tightly attached to body

IJICTHD.2021010103.f01

1.2. Gyroscope

Gyroscopes are devices that measure angular rate, and this information is processed to provide a measure of positioning.

Gyroscopes are devices mounted on a structure and able to sense an angular velocity if the structure is rotating. Many types of gyroscopes are available in market, depending on the operating physical principle and the involved technology. Most modern smartphones include Gyroscope along with Accelerometer device. Gyroscopes can be used alone or included in more complex systems, such as Gyrocompass, Inertial Measurement Unit, Inertial Navigation System and Attitude Heading Reference System (Passaro et al., 2017).

Gyroscope along with accelerometer is used to increase accuracy of the classification.

Complete Article List

Search this Journal:
Reset
Volume 15: 1 Issue (2024): Forthcoming, Available for Pre-Order
Volume 14: 4 Issues (2022): 1 Released, 3 Forthcoming
Volume 13: 4 Issues (2021)
Volume 12: 4 Issues (2020)
Volume 11: 4 Issues (2019)
Volume 10: 4 Issues (2018)
Volume 9: 4 Issues (2017)
Volume 8: 4 Issues (2016)
Volume 7: 4 Issues (2015)
Volume 6: 4 Issues (2014)
Volume 5: 4 Issues (2013)
Volume 4: 4 Issues (2012)
Volume 3: 4 Issues (2011)
Volume 2: 4 Issues (2010)
Volume 1: 4 Issues (2009)
View Complete Journal Contents Listing