PROJECT TITLE :

A Comparative Study on Human Activity Recognition Using Inertial Sensors in a Smartphone

ABSTRACT:

Activity recognition plays a vital role in bridging the gap between the low-level sensor data and therefore the high-level applications in ambient-assisted living systems. With the aim to get satisfactory recognition rate and adapt to various application scenarios, a variety of sensors have been exploited, among which, smartphone-embedded inertial sensors are widely applied because of its convenience, low value, and intrusiveness. In this paper, we tend to explore the facility of triaxial accelerometer and gyroscope engineered-during a smartphone in recognizing human physical activities in situations, where they're used simultaneously or separately. A unique Feature Selection approach is then proposed in order to pick a subset of discriminant options, construct an online activity recognizer with higher generalization ability, and scale back the smartphone power consumption. Experimental results on a publicly obtainable knowledge set show that the fusion of each accelerometer and gyroscope information contributes to obtain higher recognition performance than that of using single source information, and that the proposed feature selector outperforms 3 alternative comparative approaches in terms of four performance measures. Furthermore, great improvement in time performance will be achieved with an efficient feature selector, indicating the means of power saving and its applicability to real-world activity recognition.