基于混合动作捕捉技术的体育辅助训练系统

SPORTS AUXILIARY COACHING SYSTEM BASED ON HYBRID MOTION CAPTURE TECHNOLOGY

闫伟超

11849019

硕士

机械工程领域工程

杨再跃

2020-06-04

2020-07-20

哈尔滨工业大学

深圳

随着我国体育锻炼人数的逐年增加，体育教练资源短缺的问题愈加明显。尤其是在中西部地区，基础体育教学尚未完全普及。动作捕捉作为一种数字化记录运动物体位置和姿态的手段，为体育运动的数字化提供了技术基础。基于动作捕捉技术的数字化体育训练有望减少训练过程中对教练员的经验依赖，实现更加普遍的体育基础教育。本文研究了适用于绝大多数运动项目的体育辅助训练问题，主要创新点如下：本文首先在归纳传统体育训练模式的基础上，提出了基于知识的通用体育辅助训练框架。框架包含领域知识、受训者、运动评价和控制器四个主要模块。领域知识模块是对该项运动的练习内容、改进指令、技能指标等知识的数字化表示；受训者模块是受训者对练习内容和改进指令的动作反应；运动评价模块使用动作捕捉技术获得受训者的原始运动数据，并进一步计算得到运动特征；控制器模块依据运动评价的结果，对受训者提出改进指令或使其练习新的内容，直到受训者达到预期目标。之后，研究了通用体育训练框架在具体运动中的实施方法，即，领域知识的表达和动作捕捉系统的搭建。本文乒乓球为例，描述了框架中练习内容、运动特征和技能水平的具体形式，明确了适合于乒乓球的动作捕捉方式及运动特征的计算方法。同时，为进一步提高惯性动作捕捉系统对受训者位姿的还原精度，本文从双脚接触判断和基于接触位置约束的位姿修正两方面对原有的惯性动作捕捉系统进行了改进。仿真结果显示，修正后的人体位姿具有良好的动作平滑性。最后，开展了通用框架在乒乓球领域的应用示范。本文搭建了基于混合动作捕捉技术的乒乓球辅助训练系统，并邀请不同水平的乒乓球爱好者参与实验。实验表明：针对同一套练习内容下的不同受训者，系统计算得到的受训者水平（回球质量和动作标准性）与实际情况一致。实验结果初步验证了搭建系统的有效性，为今后进一步研究高精度、交互式的体育辅助训练系统奠定了基础。

As the number of people participating physical exercises in China increases year by year, the shortage of sports coach resources becomes more and more severe. Especially in the central and western regions, basic physical education has not yet been fully popularized. As a means of digitally recording the position and posture of moving objects, motion capture provides a technical foundation for the digitization of sports. Digital sports training based on motion capture technology is expected to reduce the experience dependence on coaches during the training process and popularize the basic physical education. This work studies the auxiliary coaching systems applicable to most sports. The main innovations of this work are as follows: Based on the refinement of traditional sports training patterns, this paper first proposes a general knowledge-based sports auxiliary coaching framework. The framework mainly includes four modules: domain knowledge, trainees, performance evaluation, and controller. The domain knowledge module digitalizes the knowledge of the sport such as practice contents, improvement instructions, and skill indicators. The trainee module stores the trainee’s action response to the practice content and improvement instructions. The performance evaluation module obtains the trainee’s original motion data from the motion capture device and further calculates the sports features. Based on the performance evaluation results, the controller decides to give the trainee improvement instructions or to practice new content. The system will operate until the trainee reaches the predefined goal. Then the implementation of the general coaching framework in specific sports is studied, that is, the expression of domain knowledge and the construction of a motion capture system. This paper takes table tennis as an example to describe the practice content, sports features, and skill level in the framework in detail, and determines the motion capture devices suitable for table tennis and features calculation methods. At the same time, in order to further improve the accuracy of the inertial motion capture system to restore the trainee’s posture, this work improves the original inertial motion capture system from two aspects of contact judgment and posture correction based on contact position constraints. The simulation results show that the corrected human posture has better smoothness of movement. Finally, a demonstration of the application of the general framework in the field of table tennis has been carried out. This work builds a table tennis auxiliary coaching system based on hybrid motion capture technology and invites table tennis players with di erent levels to participate in the experiment. The experiment shows that for di erent trainees under the same set of practice content, the trainee level (return-ball quality and standard of movement) calculated by the system is consistent with the actual situation. The experiment result has preliminarily verified the e ectiveness of the system and laid the foundation for further research on high-precision, interactive sports auxiliary coaching systems in the future.

体育辅助训练系统 动作捕捉 人机交互 运动评价 乒乓球

sports auxiliary coaching system motion capture human-computer interaction performance evaluation table tennis

中文

联合培养

南方科技大学

闫伟超. 基于混合动作捕捉技术的体育辅助训练系统[D]. 深圳. 哈尔滨工业大学,2020.