基于动态时间规整算法的健美操动作分析及评价系统设计研究  被引量:2

Research on aerobics Movement Analysis and Evaluation System Design Based on dynamic time warping Algorithm

在线阅读下载全文

作  者:常琪 CHANG Qi(Shanxi University of Chinese Medicine,Xianyang Shanxi 712046,China)

机构地区:[1]陕西中医药大学,陕西咸阳712046

出  处:《自动化与仪器仪表》2024年第2期136-139,144,共5页Automation & Instrumentation

基  金:陕西省体育局常规课题《关于老年群体体育健身服务需求的研究》(2022102)。

摘  要:随着健美操运动的发展,健美操动作的分析和评价技术成为了行业的关注点。研究通过将深度相机和动态时间规整算法进行结合,设计出了基于动态时间规整算法的健美操动作分析及评价系统。实验结果表明,在进行数据滤波测试时,研究方法生成的曲线较为平滑且没有出现与原始数据超过0.05 m的差异;在进行动作分割测试时,研究方法在静态动作中只出现3次误差,且都仅为1帧;在应用测试中,研究方法在进行关节角度相似度测试时只出现1个超过1度差异的动作;在进行动作中心时间相似度测试时只出现1次与标准值超过0.3 s的差异。说明研究方法能有效进行健美操动作分析及评价,且具有较好性能,能为健美操评价提供可行的参考方案。With the development of aerobics,the analysis and evaluation technology of aerobics movements has become the focus of the industry.Through the combination of depth camera and dynamic time warping algorithm,a aerobics action analysis and evaluation system based on dynamic time warping algorithm is designed.The experimental results indicate that during the data filtering test,the curve generated by the research method is relatively smooth and does not show any differences exceeding 0.05m from the original data;When conducting action segmentation testing,the research method only experienced 3 errors in static actions,all of which were only 1 frame;In application testing,the research method only showed one movement with a difference of more than 1 degree in joint angle similarity testing;During the action center time similarity test,there was only one difference exceeding 0.3 seconds from the standard value.It shows that the research method can effectively carry out aerobics action analysis and evaluation,and has good performance,which can provide a feasible reference scheme for aerobics evaluation.

关 键 词:健美操 动态时间规划算法 深度相机 均值滤波 动作分析 

分 类 号:TP39[自动化与计算机技术—计算机应用技术] G831.3[自动化与计算机技术—计算机科学与技术]

 

参考文献:

正在载入数据...

 

二级参考文献:

正在载入数据...

 

耦合文献:

正在载入数据...

 

引证文献:

正在载入数据...

 

二级引证文献:

正在载入数据...

 

同被引文献:

正在载入数据...

 

相关期刊文献:

正在载入数据...

相关的主题
相关的作者对象
相关的机构对象