Detection algorithm of regional peak motion based on acceleration sensor

Abstract

In this paper, an algorithm of regional peak motion detection based on acceleration sensor is designed. Firstly, the acceleration model is processed by triaxial integration method, and then the signal is filtered by Gaussian filter. Finally, the step result is obtained by combining regional peak detection, threshold limit and other methods. Through the analysis of the data collected by the acceleration sensor, we can extract the parameters representing the motion attitude of the object from the single-chip microcomputer system, and dynamically modify the parameter value set by the system according to the current system motion attitude, and compare the value with the current motion attitude parameter, so as to judge whether the current count is effective. At the same time, this device can be well applied in the industrial system to achieve convenient and effective counting function. The results show that the step counting algorithm in this paper has high accuracy.