Design of Motion Capture Methods of Cheerleaders Based on Posture Estimation and Depth Image
To effectively capture and identify cheerleaders'movements,the experiment suggests a method of capturing cheerleaders'motion based on pose estimation and depth images(PE-DI).This method is to collect the movement data of cheerleaders by a depth camera,enhance the action reconstruction effect in three-dimensional space through depth images,and then uses a posture estimation algorithm to analyze the accuracy and stability of cheerleaders'movements via a posture estimation algorithm.The results show that in the comparison of motion recognition rates,the average recognition rates of the PE-DI algorithm,NN-BV algorithm,SPR-DL algorithm and Multi-Scale Feature algorithm in five experiments were 99.24%,97.18%,94.16%and 90.21%.Finally,when the PE-DI algorithm was run,the four movements of cheerleaders,jumping,bowing and turning were all captured and divided into the correct areas,which means the algorithm is significantly better than others.The above results all show that the use of PE-DI algorithm can more precisely capture the movements of cheerlers,thus providing effective technical support for the training and performance of cheerleaders.
NETL
NSTL
万方数据
