A fast classification algorithm of walking states based on acceleration signal

This paper proposes a fast classification algorithm of walking states based on acceleration signal. The improved and simplified recursive average filtering algorithm for acceleration data is adopted and then the mean and variance of the vertical and the forward direction acceleration is extracted. After analyzing the characteristics of different walking states, the fast classification algorithm of walking states is presented. The algorithm has many advantages, such as low computational complexity, fast execution speed and low requirement for hardware. It shows good recognition rate by the experiment results.