Step Cycle Detection of Human Gait Based on Inertial Sensor Signal

As a biologic character Human gait is very important for identity recognized, health evaluation, medical monitoring. Gait cycle is one of the most basic parameters in gait analysis. We can easily calculate the gait uniformity, gait symmetry, gait continuity and other parameters based on this parameter. Especially for many of diseases estimation, such as Parkinsons disease, to get the phase synchronization, the precise time of every step event must be determined. We can simply get gait counts from traditional pedometer, but we can not get the precise step interval. In this paper, based on Pan-Tompkins algorithm used in ECG (electrocardiogram) signal, we develop one method using peak detection based on feet acceleration and angular velocity. Experimental results show the method has high precision and less error.