首页 | 本学科首页   官方微博 | 高级检索  
     检索      

基于双通道目标运动检测的乒乓球运动决策系统*
引用本文:郭〓晗.基于双通道目标运动检测的乒乓球运动决策系统*[J].广州体育学院学报,2014,34(6):67-70.
作者姓名:郭〓晗
作者单位:齐齐哈尔大学体育学院,黑龙江齐齐哈尔,161006
基金项目:黑龙江省高等学校教改工程项目
摘    要:动作是体育运动的关键技术,要规范运动员动作,不但使他们的运动成绩能够达到或接近最高水平,同时也可以避免运动员积累型运动损伤,达到这样的训练目的就需要科学的方法。目前,我国已经建立了标准动作多媒体视频数据库,利用了一种双通道目标运动检测算法,结合了运动检测技术和人的行为理解方法,实现了感知技术对体育动作量化分析和体育动作模式识别。针对乒乓球快速多变特点,以及手腕动作的精细变化与球拍的较高关联度,对乒乓球和乒乓球拍建立了离散小波变换的空时模型,剔除运动模式识别中的噪声孔洞有助于提高检测的正确性。通过定性分析和实验结果,这种基于双通道混合算法的乒乓球运动决策系统能够较为准确的识别乒乓球运动中的有误动作,为及时更正动作提供了科学和有效的依据。

关 键 词:又通道  目标检测  乒乓球  决策系统
收稿时间:2014/9/10 0:00:00

Table Tennis Decision System Based on Double Channel Object Moving Detection
GUO Han.Table Tennis Decision System Based on Double Channel Object Moving Detection[J].Journal of Guangzhou Physical Education Institute,2014,34(6):67-70.
Authors:GUO Han
Abstract:Action is the key technology in sports and we must standardize athlete''s action. If we expect they get highest level or comeclose to it, and at the same time had avoided the accumulation of sports injury of athletes, the scientific method is needed. This paper proposesa dual channel target motion detection algorithm, combines with the motion detection technology and human behavior understanding method to judge the movement results by quantitative analysis and physical action sports movement pattern recognition. The table tennis has the fast changeable characteristics, and the table tennis racket and athlete''s wrist action have high correlation. So we can establish the discrete wavelet transform of the space-time model for cancelling the noise cave.Through the qualitative analysis and the experimental results, the double channel hybrid algorithm for table tennis decision system can better identify the wrong operation of the table tennis, and provides a scientific and effective basis for timely corrective action.
Keywords:object moving detection  table tennis  decision system
本文献已被 CNKI 万方数据 等数据库收录!
点击此处可从《广州体育学院学报》浏览原始摘要信息
点击此处可从《广州体育学院学报》下载免费的PDF全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号