Research on Trajectory and Modeling of Body Movement

Abstract

Body movement analysis and identification are challenging due to high-dimension data, and can be disturbed by moving speed and position of a certain person. This paper is aiming to establish a limb trajectory model to reveal the hidden information of limb activities. According to the data onto nine parts of the human body, a three-dimension view of body activities is proposed to analyze limb positions in continuing status, represented by a sequence of body movement of the data in a three- dimension space coordinate. The motion equations of the body are obtained by regression functions under normal state and active state. The experiment of ten definitions of action shows that head, elbows, wrists, knees and ankles are strongly connected when an activity is performed.