基于机器视觉技术的运动训练系统设计  

Design of sports training system based on machine vision technology

作　　者：曾友美 宋英 ZENG Youmei;SONG Ying(Wuchang Institute of Technology,Wuhan 430065,China;Wuhan University of Engineering Science,Wuhan 430200,China)

机构地区：[1]武昌工学院,湖北武汉430065 [2]武汉工程科技学院,湖北武汉430200

出　　处：《现代电子技术》2020年第5期150-154,共5页Modern Electronics Technique

基　　金：湖北省高校省级教学改革研究项目(2017489)。

摘　　要：针对当前运动训练系统的误差考虑不全面、效果差等缺陷,设计基于机器视觉技术的运动训练系统。首先,采集运动训练视频图像,得到训练关节点运动轨迹,并依据运动轨迹和误差检测方法分析运动员运动训练测量误差;然后,依据误差成绩智能测评模块评价运动员热身运动强度、运动训练反应时间、运动训练动作精度与稳定性等,并给出提升运动效果的训练建议;最后,仿真对比实验结果表明,相比其他系统,所提系统反应时间速度快,动作控制精度高,可以获得理想的运动训练误差检测效果。In view of the defects of incomprehensive consideration of factors causing errors and poor effect in the existing sports training system,a sports training system based on machine vision technology is designed.Firstly,the video images of sports training are collected to obtain the motion tracks of trained joint points,and the measurement errors of athletes′sports training are analyzed according to the motion tracks and error detection methods.Then,according to the intelligent error evaluation module,the athletes′warm-up sports intensity,reaction time,motion accuracy and stability of sports training are evaluated.In addition,some training suggestions are given to improve the sports effect.The simulation and experiment results show that,in comparison with other systems,the proposed system has faster response speed and higher motion control accuracy,and can achieve ideal error detection effect in the sports training.

关 键 词：运动训练系统 机器视觉技术 视频图像采集 运动轨迹 误差分析 智能测评 

分 类 号：TN911.73-34[电子电信—通信与信息系统] TP391[电子电信—信息与通信工程]