Research on Dynamic Tracking and Choreographic Techniques of Modern Dance Based on Computer 3D Modeling Technology

Abstract

In this paper, we propose an improved 3D modeling method based on RecurDyn dynamics Theory to address the poor efficiency of traditional modeling techniques in dance dynamics tracking. Solid manipulation of the virtual scene is realized by geometric transformation, and singular value decomposition is applied to solve the transformation. Simulation experiments were designed to explore the tracking effect. When the MS moves at a more significant speed, the larger the Doppler frequency corresponding to the dynamic tracking and choreography technique of modern dance, the more rapidly the channel changes, and the more quickly the channel correlation decreases, such as when the MS moves at a speed of 13m/s and the correlation coefficient drops to 0.1, which corresponds to the number of sampling points within 6. From the Analysis of the ablation experiments, the errors of the three models are reduced by 9.6%, 11.2%, and 5.9%, respectively. From the first qualitative test analysis, frames 1-13 should be less spatial and 14-24 should be more spatial. From the Analysis of the first qualitative test, according to the numerical value, the average moving speed of the arm in the 53rd-68th frames is the largest, and the average moving speed of the arm in the 87th-100th frames is the smallest. Therefore, the improved algorithm can be accurately applied to modern dance dynamic tracking and choreography techniques.