基于计算机视觉技术的体育运动员错误动作智能化识别

Intelligent Recognition of Athletes'Wrong Actions Based on Computer Vision Technology

体育运动员在训练和比赛过程中,难免会出现错误动作,仅仅通过人为方式判断动作是否准确将会出现较大误差。随着计算机视觉技术的不断发展和应用,将其应用到错误动作智能化识别中将有助于提高错误动作识别率。于是文章首先介绍了计算机视觉技术,能够对拍摄图片进行数字化分析,具有较强的应用性能。然后再对运动员动作特征提取进行分析,利用贝叶斯算法对错误动作进行识别,于是得到三维视觉检测模型。最后对三维视觉检测模型进行实验研究,结果表明,相比于传统的检测方式,文章研究的模型具有更好的准确性,证明该模型应用于错误动作识别中具有可行性。

In the process of training and competition,sports athletes will inevitably have erroneous movements,and there will be large errors in judging whether the movements are accurate only by artificial means.With the continu⁃ous development and application of computer vision technology,its application to the intelligent recognition of erro⁃neous actions will help to improve the recognition rate of erroneous actions.So the paper first introduces computer vi⁃sion technology,which can perform digital analysis on the captured pictures,and has strong application performance.Then analyze the athlete's movement feature extraction,use Bayesian algorithm to identify the wrong movement,and then get the three-dimensional visual detection model.Finally,an experimental study of the three-dimensional visu⁃al detection model is carried out.The results show that the model studied in this paper has better accuracy than the traditional detection method,which proves that the model is feasible in the application of wrong action recognition.