Dance pose capture and recognition based on heterogeneous sensors

Abstract

In order to capture the movement posture of dancer in real time and effectively, a dancer motion and posture capture system based on heterogeneous sensors is proposed. Starting from the data processing of heterogeneous sensors in a single node, the single attitude measurement node of the system can be used to collect the initial posture data in the process of dancer motion in real time, achieve accurate capture of dancer movements, and filter and fuse the original motion data. In order to verify the performance and accuracy of the designed dancer movements capture system, relevant experiments are designed to verify it. The experimental results show that the algorithm designed in this experiment basically solves the interference problem of magnetic field and acceleration on attitude solution, and the attitude solution accuracy can basically meet the requirements. The fitting degree of the comparison between the extracted angle of the video sequence and the output angle of the sensor can basically meet the requirements, and the angle deviation is kept within 1 degree. The system can capture the dancer movements of the whole body and the hand, and the experimental results can meet the requirements basically, which has certain reference value.