Implementation of the Human Action State Recognition Using Smartphone

Abstract

In today's society, smartphone have become part of our lives. With regard to these small devices, the function of an accelerometer sensor has been installed as essential. Therefore, it is utilized by the public to recognize humans' action state. Among these different actions, there is a difference existing in the accelerometer value, so as to raise the nine features: Signal Magnitude Area, Variation, Correlation, Kurtosis, Skewness, Amplitude, Differential Signal, Magnitude, Mean and Standard Deviation to help us for the recognition of the action state. The bigger the difference among these features, the easier to differentiate the information of unknown actions. Therefore, we can recognize people's actions such as walking, running and static states in their daily lives.