Real-time Detection Method of Human Rhythm Movement using Jerk

Abstract

We propose a method to detect human rhythm movement in real-time using jerk. In recent years, education and cultural activities such as music and dance have attracted attention as a place for people with disabilities to participate in society. In order to promote participation of persons with disabilities, it is necessary to provide support for those with disabilities. Therefore, we are conducting study on a system that detects rhythm movements in real-time by the movements of a music conductor or dance instructor and transmits them to a performer using a haptic interface device. A three-axis acceleration sensor is used for motion detection to measure a mixed acceleration. Next, the jerk is obtained by differentiating the acceleration with respect to time. The jerk has a remarkable peak at the timing of the beat of the rhythm movement. In this paper, it was clarified by experiments that real-time detection can be performed by a microcomputer by simplifying this numerical analysis part.