A Precise Step Counting Algorithm Based on Acceleration Correlation Analysis

In view that a pedometer using zero crossing peak must be worn in the specific parts of human body, a method for realizing pedometer by using the acceleration correlation analysis was proposed. The standard deviation is used to judge whether the pedestrian is walking or not. Besides, the different patterns of personnel's sport station are identified by dynamic threshold. Then, the correlation analysis is used to judge whether the pedestrian is walking or not, because of the periodic characteristics of acceleration when the pedestrian is walking. The test by the proposed method show that, compared with the traditional zero crossing peak method, the method can accurately detect the step frequency under different sport states with an accuracy of 96%~98%.